Add auto-heal pipeline command and fix multi-source draft processing

- Add `ietf auto` command: fetches, analyzes, embeds, extracts ideas, and refreshes gaps across all sources with cost-based auto-approval - Fix SourceDocument→Draft conversion in auto fetch step - Fix gap_analysis method name in auto command - Process all 270 unrated ETSI/ISO/ITU/NIST drafts (761 total, all rated) - Update web UI templates and data layer for multi-source support Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-08 18:41:42 +01:00
parent 1ec1f69bee
commit a46a01bd8c
15 changed files with 991 additions and 381 deletions
--- a/data/drafts.db
+++ b/data/drafts.db
--- a/data/reports/gaps.md
+++ b/data/reports/gaps.md
@@ -1,125 +1,133 @@
 # Gap Analysis: IETF AI/Agent Draft Landscape
-*Generated 2026-03-08 15:15 UTC — analyzing 474 drafts, 498 technical ideas*
+*Generated 2026-03-08 17:16 UTC — analyzing 756 drafts, 501 technical ideas*
 ## Overview
-This report identifies **12 gaps** — areas, problems, or technical challenges not adequately addressed by the current 474 IETF AI/agent drafts. Each gap is cross-referenced with related drafts and extracted technical ideas to show partial coverage.
+This report identifies **11 gaps** — areas, problems, or technical challenges not adequately addressed by the current 756 IETF AI/agent drafts. Each gap is cross-referenced with related drafts and extracted technical ideas to show partial coverage.
 | Severity | Count |
 |----------|------:|
 | **CRITICAL** | 3 |
-| **HIGH** | 6 |
+| **HIGH** | 5 |
 | **MEDIUM** | 3 |
 ### Safety Deficit
-Only **46** of 474 drafts address AI safety/alignment, while **150** focus on A2A protocols and **110** on autonomous operations. The ratio of capability-building to safety is roughly **5:1**.
+Only **112** of 756 drafts address AI safety/alignment, while **157** focus on A2A protocols and **124** on autonomous operations. The ratio of capability-building to safety is roughly **2:1**.
 ---
-## 1. Real-time Agent Behavior Verification
+## 1. Agent Legal Liability Framework
 | | |
 |---|---|
 | **Severity** | CRITICAL |
-| **Category** | AI safety/alignment |
+| **Category** | policy/governance |
-| **Drafts in category** | 46 |
+| **Drafts in category** | 214 |
-Current AI safety drafts focus on governance but lack technical protocols for real-time verification that agents are behaving according to their declared policies. There's no standard way to cryptographically prove agent actions match stated intentions.
+No standard addresses who is legally responsible when autonomous agents cause harm or make binding commitments. Current frameworks focus on technical risk but ignore liability assignment between agent creators, operators, and users.
-**Evidence:** Only 46 safety drafts versus 474 total, with governance focus rather than technical verification
+**Evidence:** Real-world AI agent deployments are stalled due to unclear liability chains, especially in financial and healthcare domains
 ### Related Drafts
 **Keyword matches** (drafts mentioning gap topic):
 - [draft-an-nmrg-i2icf-cits](https://datatracker.ietf.org/doc/draft-an-nmrg-i2icf-cits/) (score 3.7) — Interface to In-Network Computing Functions for Cooperative Intelligent Transpor
 - [draft-zhao-detnet-enhanced-use-cases](https://datatracker.ietf.org/doc/draft-zhao-detnet-enhanced-use-cases/) (score 3.2) — Enhanced Use Cases for Scaling Deterministic Networks
 - [draft-zhang-rvp-problem-statement](https://datatracker.ietf.org/doc/draft-zhang-rvp-problem-statement/) (score 3.5) — Problem Statements and Requirements of Real-Virtual Agent Protocol (RVP): Commun
 - [draft-yuan-rtgwg-traffic-agent-usecase](https://datatracker.ietf.org/doc/draft-yuan-rtgwg-traffic-agent-usecase/) (score 3.7) — Use cases of the AI Network Traffic Optimization Agent
 - [draft-altanai-aipref-realtime-protocol-bindings](https://datatracker.ietf.org/doc/draft-altanai-aipref-realtime-protocol-bindings/) (score 3.6) — AI Preferences for Real-Time Protocol Bindings
 - [draft-zheng-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-agent-identity-management/) (score 3.7) — Agent Identity Managenment
 - [draft-zheng-dispatch-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-dispatch-agent-identity-management/) (score 3.3) — Agent Identity Managenment
 - [draft-li-dmsc-macp](https://datatracker.ietf.org/doc/draft-li-dmsc-macp/) (score 4.2) — Multi-agent Collaboration Protocol Suite
 - [draft-fu-nmop-agent-communication-framework](https://datatracker.ietf.org/doc/draft-fu-nmop-agent-communication-framework/) (score 3.0) — Agent Communication Framework for Network AIOps
 - [draft-zyyhl-agent-networks-framework](https://datatracker.ietf.org/doc/draft-zyyhl-agent-networks-framework/) (score 3.6) — Framework for AI Agent Networks
 - [draft-madhavan-aipref-displaybasedpref](https://datatracker.ietf.org/doc/draft-madhavan-aipref-displaybasedpref/) (score 2.5) — A Vocabulary for Controlling Usage of Content Collected by Search and AI Crawler
-**Top-rated in AI safety/alignment** (46 drafts):
+**Top-rated in policy/governance** (214 drafts):
 - [draft-cowles-volt](https://datatracker.ietf.org/doc/draft-cowles-volt/) (4.8) — Defines tamper-evident execution trace format for AI agent workflows using hash chains and cryptogra
 - [draft-aylward-daap-v2](https://datatracker.ietf.org/doc/draft-aylward-daap-v2/) (4.8) — Defines comprehensive protocol for AI agent accountability including authentication, monitoring, and
 - [itu-t-y-3172](https://datatracker.ietf.org/doc/itu-t-y-3172/) (4.7) — Establishes comprehensive architectural framework for machine learning deployment in future networks
 - [iso-iec-22989-2022](https://datatracker.ietf.org/doc/iso-iec-22989-2022/) (4.7) — ISO/IEC standard defining core AI concepts and establishing standardized terminology across the fiel
 - [iso-iec-42001-2023](https://datatracker.ietf.org/doc/iso-iec-42001-2023/) (4.6) — ISO/IEC standard establishing comprehensive AI management system requirements covering governance, r
 ### Partially Addressing Ideas
 51 extracted ideas touch on this gap:
 | Idea | Draft | Type |
 |------|-------|------|
 | Structured Responsibility and Traceability Architecture | draft-takagi-srta-trinity | architecture |
 | Intelligent Agent Communication Gateway Architecture | draft-agent-gw | architecture |
 | Tiered Risk-Based Governance for Autonomous AI Agents | draft-aylward-aiga-1 | architecture |
 | Tiered Risk-Based Governance for Autonomous AI Agents | draft-aylward-aiga-2 | architecture |
 | Distributed AI Accountability Protocol | draft-aylward-daap-v2 | protocol |
 | Zero Trust Runtime Agent Architecture | draft-berlinai-vera | architecture |
 | Agentic Hypercall Protocol | draft-campbell-agentic-http | pattern |
 | Context Distribution Optimization Framework | draft-chang-agent-context-interaction | mechanism |
 *...and 43 more*
 ---
 ## 2. Agent Capability Degradation Detection
 | | |
 |---|---|
 | **Severity** | CRITICAL |
 | **Category** | AI safety/alignment |
 | **Drafts in category** | 112 |
 No standard defines how to detect when an agent's capabilities are degrading due to concept drift, adversarial inputs, or model corruption. Current monitoring focuses on system metrics not capability assessment.
 **Evidence:** Production AI systems show gradual performance degradation that goes undetected until major failures occur
 ### Related Drafts
 **Keyword matches** (drafts mentioning gap topic):
 - [draft-zheng-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-agent-identity-management/) (score 3.7) — Agent Identity Managenment
 - [draft-zheng-dispatch-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-dispatch-agent-identity-management/) (score 3.3) — Agent Identity Managenment
 - [draft-li-dmsc-macp](https://datatracker.ietf.org/doc/draft-li-dmsc-macp/) (score 4.2) — Multi-agent Collaboration Protocol Suite
 - [draft-fu-nmop-agent-communication-framework](https://datatracker.ietf.org/doc/draft-fu-nmop-agent-communication-framework/) (score 3.0) — Agent Communication Framework for Network AIOps
 - [draft-zyyhl-agent-networks-framework](https://datatracker.ietf.org/doc/draft-zyyhl-agent-networks-framework/) (score 3.6) — Framework for AI Agent Networks
 - [draft-li-dmsc-inf-architecture](https://datatracker.ietf.org/doc/draft-li-dmsc-inf-architecture/) (score 3.1) — Dynamic Multi-agent Secured Collaboration Infrastructure Architecture
 **Top-rated in AI safety/alignment** (112 drafts):
 - [draft-cowles-volt](https://datatracker.ietf.org/doc/draft-cowles-volt/) (4.8) — Defines tamper-evident execution trace format for AI agent workflows using hash chains and cryptogra
 - [draft-aylward-daap-v2](https://datatracker.ietf.org/doc/draft-aylward-daap-v2/) (4.8) — Defines comprehensive protocol for AI agent accountability including authentication, monitoring, and
 - [draft-guy-bary-stamp-protocol](https://datatracker.ietf.org/doc/draft-guy-bary-stamp-protocol/) (4.6) — Defines STAMP protocol for cryptographic delegation and proof in AI agent systems. Provides task-bou
 - [draft-drake-email-tpm-attestation](https://datatracker.ietf.org/doc/draft-drake-email-tpm-attestation/) (4.6) — Defines hardware attestation for email using TPM verification chains to prevent spam and provide Syb
- [draft-goswami-agentic-jwt](https://datatracker.ietf.org/doc/draft-goswami-agentic-jwt/) (4.5) — Extends OAuth 2.0 with Agentic JWT to address authorization challenges in autonomous AI systems. Int
+- [iso-iec-42001-2023](https://datatracker.ietf.org/doc/iso-iec-42001-2023/) (4.6) — ISO/IEC standard establishing comprehensive AI management system requirements covering governance, r
 ### Partially Addressing Ideas
-18 extracted ideas touch on this gap:
+22 extracted ideas touch on this gap:
 | Idea | Draft | Type |
 |------|-------|------|
-| Distributed AI Accountability Protocol | draft-aylward-daap-v2 | protocol |
+| Intelligent Agent Communication Gateway Architecture | draft-agent-gw | architecture |
-| AGENTS.TXT Policy File | draft-srijal-agents-policy | protocol |
+| AI-Native Network Protocol (AINP) | draft-ainp-protocol | protocol |
-| AI Network Security Agent | draft-yuan-rtgwg-security-agent-usecase | architecture |
+| Agentic Data Optimization Layer (ADOL) | draft-chang-agent-token-efficient | protocol |
-| A2A Protocol Transport over MOQT | draft-a2a-moqt-transport | protocol |
+| Structured OAuth Scope Syntax for Agent Permissions | draft-chen-oauth-scope-agent-extensions | extension |
-| Post-Discovery Authorization Handshake | draft-barney-caam | protocol |
+| Capability-based Agent Discovery Mechanism | draft-cui-ai-agent-discovery-invocation | mechanism |
-| Evidence-based Autonomy Maturity Model | draft-berlinai-vera | mechanism |
+| Intent-based Agent Selection | draft-cui-ai-agent-discovery-invocation | extension |
-| Verifiable Agent Conversation Format | draft-birkholz-verifiable-agent-conversations | protocol |
+| Agent Attachment Protocol | draft-dunbar-agent-attachment | protocol |
-| Intent-Based Just-in-Time Authorization | draft-chen-agent-decoupled-authorization-model | architecture |
+| EAT Extensions for Agent Capability Attestation | draft-huang-rats-agentic-eat-cap-attest | extension |
-*...and 10 more*
+*...and 14 more*
 ---
-## 2. Multi-Agent Consensus Under Byzantine Conditions
+## 3. Emergency Agent Override Protocols
 | | |
 |---|---|
 | **Severity** | CRITICAL |
 | **Category** | A2A protocols |
 | **Drafts in category** | 150 |
 While agent discovery and A2A protocols exist, there's no framework for handling consensus when some agents may be compromised or malicious. Critical for autonomous systems making collective decisions.
 **Evidence:** Complex autonomous systems require Byzantine fault tolerance but it's absent from protocol designs
 ### Related Drafts
 **Keyword matches** (drafts mentioning gap topic):
 - [draft-li-dmsc-mcps-agw](https://datatracker.ietf.org/doc/draft-li-dmsc-mcps-agw/) (score 3.5) — Multi-agent Collaboration Protocol Suite based on Agent Gateway
 - [draft-li-dmsc-inf-architecture](https://datatracker.ietf.org/doc/draft-li-dmsc-inf-architecture/) (score 3.1) — Dynamic Multi-agent Secured Collaboration Infrastructure Architecture
 - [draft-yue-anima-agent-recovery-networks](https://datatracker.ietf.org/doc/draft-yue-anima-agent-recovery-networks/) (score 4.1) — Task-Oriented Multi-Agent Recovery Framework for High-Reliability in Converged M
 - [draft-chang-agent-context-interaction](https://datatracker.ietf.org/doc/draft-chang-agent-context-interaction/) (score 2.9) — Agent Context Interaction Optimizations
 - [draft-fu-nmop-agent-communication-framework](https://datatracker.ietf.org/doc/draft-fu-nmop-agent-communication-framework/) (score 3.0) — Agent Communication Framework for Network AIOps
 - [draft-ramakrishna-satp-views-addresses](https://datatracker.ietf.org/doc/draft-ramakrishna-satp-views-addresses/) (score 3.4) — Views and View Addresses for Secure Asset Transfer
 **Top-rated in A2A protocols** (150 drafts):
 - [draft-guy-bary-stamp-protocol](https://datatracker.ietf.org/doc/draft-guy-bary-stamp-protocol/) (4.6) — Defines STAMP protocol for cryptographic delegation and proof in AI agent systems. Provides task-bou
 - [draft-williams-netmod-lm-hierarchy-topology](https://datatracker.ietf.org/doc/draft-williams-netmod-lm-hierarchy-topology/) (4.6) — Defines YANG data model for hierarchical language model coordination across tiny, small, and large L
 - [draft-ietf-lake-edhoc](https://datatracker.ietf.org/doc/draft-ietf-lake-edhoc/) (4.6) — Specifies EDHOC, a compact authenticated Diffie-Hellman key exchange protocol for constrained enviro
 - [draft-chang-agent-token-efficient](https://datatracker.ietf.org/doc/draft-chang-agent-token-efficient/) (4.5) — Defines ADOL (Agentic Data Optimization Layer) to address token bloat in agent communication protoco
 - [draft-chen-oauth-rar-agent-extensions](https://datatracker.ietf.org/doc/draft-chen-oauth-rar-agent-extensions/) (4.2) — Extends OAuth RAR with policy_context and lifecycle_binding members for AI agent environments. Enabl
 ### Partially Addressing Ideas
 2 extracted ideas touch on this gap:
 | Idea | Draft | Type |
 |------|-------|------|
 | ASRank Structural Vulnerability Analysis | draft-xu-sidrops-asrank-vulnerabilities | requirement |
 | MCP and A2A Complementary Solutions for Network Management | draft-zeng-opsawg-applicability-mcp-a2a | architecture |
 ---
 ## 3. Emergency Agent Shutdown Coordination
 | | |
 |---|---|
 | **Severity** | CRITICAL |
 | **Category** | AI safety/alignment |
-| **Drafts in category** | 46 |
+| **Drafts in category** | 112 |
-Missing protocols for coordinated emergency shutdown of autonomous agent networks when safety issues are detected. Individual agent controls exist but not network-wide coordination mechanisms.
+No standard defines how to safely emergency-stop or override autonomous agents across distributed systems when they exhibit dangerous behavior. Current approaches assume centralized control that may not exist.
-**Evidence:** Human-in-the-loop drafts exist but no emergency coordination protocols for autonomous systems
+**Evidence:** Incidents with autonomous trading systems and industrial controls show need for fail-safe override mechanisms
 ### Related Drafts
@@ -128,177 +136,48 @@ Missing protocols for coordinated emergency shutdown of autonomous agent network
 - [draft-aylward-daap-v2](https://datatracker.ietf.org/doc/draft-aylward-daap-v2/) (score 4.8) — Distributed AI Accountability Protocol (DAAP) Version 2.0
 - [draft-khatri-sipcore-call-transfer-fail-response](https://datatracker.ietf.org/doc/draft-khatri-sipcore-call-transfer-fail-response/) (score 3.3) — A SIP Response Code (497) for Call Transfer Failure
 - [draft-cui-dmsc-agent-cdi](https://datatracker.ietf.org/doc/draft-cui-dmsc-agent-cdi/) (score 3.0) — Cross-Domain Interoperability Framework for AI Agent Collaboration
- [draft-yu-ai-agent-use-cases-in-6g](https://datatracker.ietf.org/doc/draft-yu-ai-agent-use-cases-in-6g/) (score 2.5) — AI Agent Use Cases and Requirements in 6G Network
+- [iso-22733-1-2021](https://datatracker.ietf.org/doc/iso-22733-1-2021/) (score 4.3) — ISO 22733-1:2021: Road vehicles — Test method to evaluate the performance of aut
- [draft-zhang-rvp-problem-statement](https://datatracker.ietf.org/doc/draft-zhang-rvp-problem-statement/) (score 3.5) — Problem Statements and Requirements of Real-Virtual Agent Protocol (RVP): Commun
+- [iso-22733-1-2022](https://datatracker.ietf.org/doc/iso-22733-1-2022/) (score 2.5) — ISO 22733-1:2022: Road vehicles — Test method to evaluate the performance of aut
 - [draft-zheng-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-agent-identity-management/) (score 3.7) — Agent Identity Managenment
-**Top-rated in AI safety/alignment** (46 drafts):
+**Top-rated in AI safety/alignment** (112 drafts):
 - [draft-cowles-volt](https://datatracker.ietf.org/doc/draft-cowles-volt/) (4.8) — Defines tamper-evident execution trace format for AI agent workflows using hash chains and cryptogra
 - [draft-aylward-daap-v2](https://datatracker.ietf.org/doc/draft-aylward-daap-v2/) (4.8) — Defines comprehensive protocol for AI agent accountability including authentication, monitoring, and
 - [draft-guy-bary-stamp-protocol](https://datatracker.ietf.org/doc/draft-guy-bary-stamp-protocol/) (4.6) — Defines STAMP protocol for cryptographic delegation and proof in AI agent systems. Provides task-bou
 - [draft-drake-email-tpm-attestation](https://datatracker.ietf.org/doc/draft-drake-email-tpm-attestation/) (4.6) — Defines hardware attestation for email using TPM verification chains to prevent spam and provide Syb
- [draft-goswami-agentic-jwt](https://datatracker.ietf.org/doc/draft-goswami-agentic-jwt/) (4.5) — Extends OAuth 2.0 with Agentic JWT to address authorization challenges in autonomous AI systems. Int
+- [iso-iec-42001-2023](https://datatracker.ietf.org/doc/iso-iec-42001-2023/) (4.6) — ISO/IEC standard establishing comprehensive AI management system requirements covering governance, r
 ### Partially Addressing Ideas
-9 extracted ideas touch on this gap:
+15 extracted ideas touch on this gap:
 | Idea | Draft | Type |
 |------|-------|------|
-| Distributed AI Accountability Protocol | draft-aylward-daap-v2 | protocol |
+| LLM-Enhanced Autonomic Service Agent Architecture | draft-han-anima-ai-asa | architecture |
-| Agentic network architecture for multi-agent coordination | draft-chuyi-nmrg-agentic-network-inference | architecture |
+| Multipath Traffic Engineering Capabilities Advertisement | draft-kompella-lsr-mptecap | mechanism |
-| Dynamic Task Coordination Requirements for AI Agents | draft-cui-ai-agent-task | requirement |
+| Agent Collaboration Protocols Architecture | draft-liu-dmsc-acps-arc | architecture |
-| Multi-Agent Communication Framework for AIOps | draft-fu-nmop-agent-communication-framework | architecture |
+| Agent Lifecycle Support | draft-liu-dmsc-acps-arc | protocol |
-| Meta-Layer Coordination Substrate | draft-meta-layer-overview | architecture |
+| Zero Trust Interoperability Framework | draft-liu-saag-zt-problem-statement | requirement |
-| Trinity Configuration for Agent Coordination | draft-takagi-srta-trinity | pattern |
+| Cross-device Communication Protocol Gap Analysis for Network AI Agents | draft-mao-rtgwg-agent-comm-protocol-gap-analysis | requirement |
-| Internet of Agents Task Protocol for heterogeneous collaboration | draft-yang-dmsc-ioa-task-protocol | protocol |
+| Comparative analysis of messaging protocols for agentic AI | draft-mpsb-agntcy-messaging | pattern |
-| Task-Oriented Multi-Agent Recovery Framework | draft-yue-anima-agent-recovery-networks | architecture |
+| Cross-Device Communication Framework for Network AI Agents | draft-mzsg-rtgwg-agent-cross-device-comm-framework | architecture |
-*...and 1 more*
+*...and 7 more*
 ---
-## 4. Cross-Protocol Agent Migration
+## 4. Cross-Domain Agent Identity Portability
 | | |
 |---|---|
 | **Severity** | HIGH |
-| **Category** | A2A protocols |
+| **Category** | agent identity/auth |
-| **Drafts in category** | 150 |
+| **Drafts in category** | 160 |
-While A2A protocols exist, there's no standardized mechanism for agents to migrate between different protocol frameworks or service providers while maintaining state and identity. This creates vendor lock-in and limits agent portability across heterogeneous systems.
+Agents cannot maintain consistent identity across different organizational domains or standards bodies' protocols. IETF defines authentication within network boundaries while ISO focuses on domain-specific identity, creating fragmentation.
-**Evidence:** 150 A2A protocol drafts with high overlap suggest fragmentation without migration solutions
+**Evidence:** Enterprise deployments require agents to work across cloud providers, on-premises systems, and partner networks with different identity systems
 ### Related Drafts
 **Keyword matches** (drafts mentioning gap topic):
 - [draft-cowles-volt](https://datatracker.ietf.org/doc/draft-cowles-volt/) (score 4.8) — Verifiable Operations Ledger and Trace (VOLT) Protocol
 - [draft-han-ai-agent-impact-infra](https://datatracker.ietf.org/doc/draft-han-ai-agent-impact-infra/) (score 2.3) — The Impact of AI Agent to Network Infrastructure
 - [draft-narajala-ans](https://datatracker.ietf.org/doc/draft-narajala-ans/) (score 4.2) — Agent Name Service (ANS): A Universal Directory for Secure AI Agent Discovery an
 - [draft-ietf-emu-eap-edhoc](https://datatracker.ietf.org/doc/draft-ietf-emu-eap-edhoc/) (score 3.2) — Using the Extensible Authentication Protocol (EAP) with Ephemeral Diffie-Hellman
 - [draft-howe-sipcore-mcp-extension](https://datatracker.ietf.org/doc/draft-howe-sipcore-mcp-extension/) (score 3.7) — SIP Extension for Model Context Protocol (MCP)
 - [draft-zheng-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-agent-identity-management/) (score 3.7) — Agent Identity Managenment
 **Top-rated in A2A protocols** (150 drafts):
 - [draft-guy-bary-stamp-protocol](https://datatracker.ietf.org/doc/draft-guy-bary-stamp-protocol/) (4.6) — Defines STAMP protocol for cryptographic delegation and proof in AI agent systems. Provides task-bou
 - [draft-williams-netmod-lm-hierarchy-topology](https://datatracker.ietf.org/doc/draft-williams-netmod-lm-hierarchy-topology/) (4.6) — Defines YANG data model for hierarchical language model coordination across tiny, small, and large L
 - [draft-ietf-lake-edhoc](https://datatracker.ietf.org/doc/draft-ietf-lake-edhoc/) (4.6) — Specifies EDHOC, a compact authenticated Diffie-Hellman key exchange protocol for constrained enviro
 - [draft-chang-agent-token-efficient](https://datatracker.ietf.org/doc/draft-chang-agent-token-efficient/) (4.5) — Defines ADOL (Agentic Data Optimization Layer) to address token bloat in agent communication protoco
 - [draft-chen-oauth-rar-agent-extensions](https://datatracker.ietf.org/doc/draft-chen-oauth-rar-agent-extensions/) (4.2) — Extends OAuth RAR with policy_context and lifecycle_binding members for AI agent environments. Enabl
 ### Partially Addressing Ideas
 No directly related technical ideas found in current drafts — this gap is entirely unaddressed.
 ---
 ## 5. Agent Resource Accounting and Billing
 | | |
 |---|---|
 | **Severity** | HIGH |
 | **Category** | new |
 | **Drafts in category** | 0 |
 No standardized protocols exist for tracking and billing computational resources consumed by autonomous agents across distributed systems. This is essential for commercial deployment but completely unaddressed.
 **Evidence:** High focus on protocols and deployment but zero drafts addressing economic models
 ### Related Drafts
 **Keyword matches** (drafts mentioning gap topic):
 - [draft-zheng-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-agent-identity-management/) (score 3.7) — Agent Identity Managenment
 - [draft-li-dmsc-macp](https://datatracker.ietf.org/doc/draft-li-dmsc-macp/) (score 4.2) — Multi-agent Collaboration Protocol Suite
 - [draft-zheng-dispatch-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-dispatch-agent-identity-management/) (score 3.3) — Agent Identity Managenment
 - [draft-fu-nmop-agent-communication-framework](https://datatracker.ietf.org/doc/draft-fu-nmop-agent-communication-framework/) (score 3.0) — Agent Communication Framework for Network AIOps
 - [draft-zyyhl-agent-networks-framework](https://datatracker.ietf.org/doc/draft-zyyhl-agent-networks-framework/) (score 3.6) — Framework for AI Agent Networks
 - [draft-jia-oauth-scope-aggregation](https://datatracker.ietf.org/doc/draft-jia-oauth-scope-aggregation/) (score 3.5) — OAuth 2.0 Scope Aggregation for Multi-Step AI Agent Workflows
 ### Partially Addressing Ideas
 10 extracted ideas touch on this gap:
 | Idea | Draft | Type |
 |------|-------|------|
 | SCIM 2.0 Extension for Agents and Agentic Applications | draft-abbey-scim-agent-extension | extension |
 | Events Query Protocol | draft-gupta-httpapi-events-query | protocol |
 | Micro Agent Communication Protocol (µACP) | draft-mallick-muacp | protocol |
 | MOQT Binding for A2A and MCP Protocols | draft-nandakumar-ai-agent-moq-transport | extension |
 | AI Agent Protocol Requirements | draft-rosenberg-ai-protocols | requirement |
 | SCIM 2.0 Agent Extension | draft-scim-agent-extension | extension |
 | Authorized Connection Policy Framework | draft-steckbeck-ua-conn-sec | mechanism |
 | Agent Workflow Protocol Well-Known Resource | draft-vinaysingh-awp-wellknown | extension |
 *...and 2 more*
 ---
 ## 6. Agent Capability Advertisement Verification
 | | |
 |---|---|
 | **Severity** | HIGH |
 | **Category** | Agent discovery/reg |
 | **Drafts in category** | 87 |
 While agent discovery protocols exist, there's no way to cryptographically verify that advertised agent capabilities are accurate. Agents could falsely claim capabilities leading to system failures.
 **Evidence:** 87 discovery drafts but no mention of capability verification mechanisms
 ### Related Drafts
 **Keyword matches** (drafts mentioning gap topic):
 - [draft-zheng-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-agent-identity-management/) (score 3.7) — Agent Identity Managenment
 - [draft-li-dmsc-macp](https://datatracker.ietf.org/doc/draft-li-dmsc-macp/) (score 4.2) — Multi-agent Collaboration Protocol Suite
 - [draft-zheng-dispatch-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-dispatch-agent-identity-management/) (score 3.3) — Agent Identity Managenment
 - [draft-fu-nmop-agent-communication-framework](https://datatracker.ietf.org/doc/draft-fu-nmop-agent-communication-framework/) (score 3.0) — Agent Communication Framework for Network AIOps
 - [draft-zyyhl-agent-networks-framework](https://datatracker.ietf.org/doc/draft-zyyhl-agent-networks-framework/) (score 3.6) — Framework for AI Agent Networks
 - [draft-li-dmsc-inf-architecture](https://datatracker.ietf.org/doc/draft-li-dmsc-inf-architecture/) (score 3.1) — Dynamic Multi-agent Secured Collaboration Infrastructure Architecture
 **Top-rated in Agent discovery/reg** (87 drafts):
 - [draft-narajala-ans](https://datatracker.ietf.org/doc/draft-narajala-ans/) (4.2) — Introduces Agent Name Service (ANS) as a DNS-based universal directory for AI agent discovery and ve
 - [draft-li-dmsc-macp](https://datatracker.ietf.org/doc/draft-li-dmsc-macp/) (4.2) — Specifies a comprehensive multi-agent collaboration protocol suite using Agent Gateways for registra
 - [draft-cui-dns-native-agent-naming-resolution](https://datatracker.ietf.org/doc/draft-cui-dns-native-agent-naming-resolution/) (4.1) — Specifies DNS-native naming and resolution for AI agents using FQDNs and SVCB records. Emphasizes DN
 - [draft-nederveld-adl](https://datatracker.ietf.org/doc/draft-nederveld-adl/) (4.1) — Defines ADL, a JSON-based standard for describing AI agents including their capabilities, tools, per
 - [draft-rosenberg-ai-protocols](https://datatracker.ietf.org/doc/draft-rosenberg-ai-protocols/) (4.1) — Establishes framework for AI agent communications on the Internet, surveying existing protocols like
 ### Partially Addressing Ideas
 27 extracted ideas touch on this gap:
 | Idea | Draft | Type |
 |------|-------|------|
 | DNS-based AI Agent Discovery | draft-mozleywilliams-dnsop-bandaid | mechanism |
 | DNS namespace for AI agent discovery | draft-mozleywilliams-dnsop-dnsaid | mechanism |
 | Agent Registration and Discovery Protocol | draft-pioli-agent-discovery | protocol |
 | Intent-based Agent Interconnection Protocol | draft-sun-zhang-iaip | protocol |
 | Capability Advertisement and Intent Resolution | draft-sz-dmsc-iaip | mechanism |
 | Intelligent Agent Communication Gateway Architecture | draft-agent-gw | architecture |
 | AI-Native Network Protocol (AINP) | draft-ainp-protocol | protocol |
 | Distributed AI Accountability Protocol | draft-aylward-daap-v2 | protocol |
 *...and 19 more*
 ---
 ## 7. Cross-Domain Agent Communication Security
 | | |
 |---|---|
 | **Severity** | HIGH |
 | **Category** | Agent identity/auth |
 | **Drafts in category** | 145 |
 Current identity/auth solutions don't address secure communication between agents operating in different security domains or trust boundaries. Critical for enterprise and government deployments.
 **Evidence:** 145 identity drafts show awareness but cross-domain scenarios appear unaddressed
 ### Related Drafts
@@ -311,7 +190,7 @@ Current identity/auth solutions don't address secure communication between agent
 - [draft-intellinode-ai-semantic-contract](https://datatracker.ietf.org/doc/draft-intellinode-ai-semantic-contract/) (score 3.2) — Semantic-Driven Traffic Shaping Contract for AI Networks
 - [draft-zheng-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-agent-identity-management/) (score 3.7) — Agent Identity Managenment
-**Top-rated in Agent identity/auth** (145 drafts):
+**Top-rated in agent identity/auth** (160 drafts):
 - [draft-cowles-volt](https://datatracker.ietf.org/doc/draft-cowles-volt/) (4.8) — Defines tamper-evident execution trace format for AI agent workflows using hash chains and cryptogra
 - [draft-aylward-daap-v2](https://datatracker.ietf.org/doc/draft-aylward-daap-v2/) (4.8) — Defines comprehensive protocol for AI agent accountability including authentication, monitoring, and
@@ -321,90 +200,103 @@ Current identity/auth solutions don't address secure communication between agent
 ### Partially Addressing Ideas
-54 extracted ideas touch on this gap:
+33 extracted ideas touch on this gap:
 | Idea | Draft | Type |
 |------|-------|------|
-| Agent Gateway Intercommunication Framework | draft-han-rtgwg-agent-gateway-intercomm-framework | architecture |
+| Cross-Domain Agent Interoperability Framework | draft-cui-dmsc-agent-cdi | architecture |
-| Agent Gateway Requirements | draft-liu-rtgwg-agent-gateway-requirements | requirement |
+| Cross-Domain Authorization Information Sharing for Multi-Agent Systems | draft-diaconu-agents-authz-info-sharing | mechanism |
-| AI Agent Security Requirements Framework | draft-ni-a2a-ai-agent-security-requirements | requirement |
+| Agent Authorization Profile for OAuth 2.0 | draft-aap-oauth-profile | extension |
-| Centralized Gateway for Multi-Agent Communication | draft-song-dmsc-problem-statement | architecture |
+| SCIM 2.0 Extension for Agents and Agentic Applications | draft-abbey-scim-agent-extension | extension |
-| Multi-Tenant Policy Enforcement Infrastructure | draft-song-dmsc-problem-statement | architecture |
+| Distributed AI Accountability Protocol | draft-aylward-daap-v2 | protocol |
-| Intelligent Agent Communication Gateway Architecture | draft-agent-gw | architecture |
+| Intent-Based Just-in-Time Authorization | draft-chen-agent-decoupled-authorization-model | architecture |
-| AI-Native Network Protocol (AINP) | draft-ainp-protocol | protocol |
+| Dynamic Behavior-Based Authentication and Authorization Requirements | draft-chen-ai-agent-auth-new-requirements | requirement |
-| Agent-to-Agent Communication in Transportation Networks | draft-an-nmrg-i2icf-cits | pattern |
+| Agentic network architecture for multi-agent coordination | draft-chuyi-nmrg-agentic-network-inference | architecture |
-*...and 46 more*
+*...and 25 more*
 ---
-## 8. Agent Performance Degradation Detection
+## 5. Real-Time Agent Behavior Explanation
 | | |
 |---|---|
 | **Severity** | HIGH |
-| **Category** | new |
+| **Category** | human-agent interaction |
-| **Drafts in category** | 0 |
+| **Drafts in category** | 57 |
-No standardized protocols exist for detecting when AI agents are experiencing model drift, adversarial attacks, or performance degradation. Essential for maintaining autonomous system reliability.
+No standard defines how autonomous agents should provide real-time explanations of their decision-making to humans during operation. Current explainable AI frameworks are post-hoc rather than interactive.
-**Evidence:** ML traffic management exists but not agent health monitoring standards
+**Evidence:** Regulatory requirements emerging in EU AI Act and similar legislation demand real-time explainability for high-risk AI systems
 ### Related Drafts
 **Keyword matches** (drafts mentioning gap topic):
 - [draft-an-nmrg-i2icf-cits](https://datatracker.ietf.org/doc/draft-an-nmrg-i2icf-cits/) (score 3.7) — Interface to In-Network Computing Functions for Cooperative Intelligent Transpor
 - [draft-zhao-detnet-enhanced-use-cases](https://datatracker.ietf.org/doc/draft-zhao-detnet-enhanced-use-cases/) (score 3.2) — Enhanced Use Cases for Scaling Deterministic Networks
 - [draft-zhang-rvp-problem-statement](https://datatracker.ietf.org/doc/draft-zhang-rvp-problem-statement/) (score 3.5) — Problem Statements and Requirements of Real-Virtual Agent Protocol (RVP): Commun
 - [draft-yuan-rtgwg-traffic-agent-usecase](https://datatracker.ietf.org/doc/draft-yuan-rtgwg-traffic-agent-usecase/) (score 3.7) — Use cases of the AI Network Traffic Optimization Agent
 - [draft-altanai-aipref-realtime-protocol-bindings](https://datatracker.ietf.org/doc/draft-altanai-aipref-realtime-protocol-bindings/) (score 3.6) — AI Preferences for Real-Time Protocol Bindings
 - [draft-zheng-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-agent-identity-management/) (score 3.7) — Agent Identity Managenment
- [draft-li-dmsc-macp](https://datatracker.ietf.org/doc/draft-li-dmsc-macp/) (score 4.2) — Multi-agent Collaboration Protocol Suite
+
- [draft-zheng-dispatch-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-dispatch-agent-identity-management/) (score 3.3) — Agent Identity Managenment
+**Top-rated in human-agent interaction** (57 drafts):
- [draft-fu-nmop-agent-communication-framework](https://datatracker.ietf.org/doc/draft-fu-nmop-agent-communication-framework/) (score 3.0) — Agent Communication Framework for Network AIOps
+
- [draft-zyyhl-agent-networks-framework](https://datatracker.ietf.org/doc/draft-zyyhl-agent-networks-framework/) (score 3.6) — Framework for AI Agent Networks
+- [draft-drake-email-tpm-attestation](https://datatracker.ietf.org/doc/draft-drake-email-tpm-attestation/) (4.6) — Defines hardware attestation for email using TPM verification chains to prevent spam and provide Syb
- [draft-xiong-rtgwg-use-cases-hp-wan](https://datatracker.ietf.org/doc/draft-xiong-rtgwg-use-cases-hp-wan/) (score 2.6) — Use Cases for High-performance Wide Area Network
+- [iso-37181-2022](https://datatracker.ietf.org/doc/iso-37181-2022/) (4.5) — Establishes guidelines for introducing and organizing autonomous vehicles on public roads. Addresses
 - [iso-pas-8800-2024](https://datatracker.ietf.org/doc/iso-pas-8800-2024/) (4.5) — Addresses safety-related E/E systems using AI technology in series-production road vehicles, coverin
 - [draft-ietf-aipref-vocab](https://datatracker.ietf.org/doc/draft-ietf-aipref-vocab/) (4.4) — Defines a standardized vocabulary for expressing preferences about how digital assets should be used
 - [iso-iec-ts-6254-2025](https://datatracker.ietf.org/doc/iso-iec-ts-6254-2025/) (4.4) — Provides approaches for achieving explainability and interpretability of ML and AI systems across li
 ### Partially Addressing Ideas
-5 extracted ideas touch on this gap:
+13 extracted ideas touch on this gap:
 | Idea | Draft | Type |
 |------|-------|------|
 | Virtual In-Cloud Router as IPv6 Enhancement Agent | draft-he-yi-srv6ops-ipv6-enhancemnet-in-cloud-uc | architecture |
 | 6G Agent Protocol Requirements and Enabling Technologies | draft-hw-ai-agent-6g | requirement |
 | Comparative analysis of messaging protocols for agentic AI | draft-mpsb-agntcy-messaging | pattern |
 | AI Network Security Agent | draft-yuan-rtgwg-security-agent-usecase | architecture |
-| Task-Oriented Multi-Agent Recovery Framework | draft-yue-anima-agent-recovery-networks | architecture |
+| A2A Protocol Transport over MOQT | draft-a2a-moqt-transport | protocol |
 | Distributed AI Accountability Protocol | draft-aylward-daap-v2 | protocol |
 | Post-Discovery Authorization Handshake | draft-barney-caam | protocol |
 | Evidence-based Autonomy Maturity Model | draft-berlinai-vera | mechanism |
 | Intent-Based Just-in-Time Authorization | draft-chen-agent-decoupled-authorization-model | architecture |
 | Dynamic Behavior-Based Authentication and Authorization Requirements | draft-chen-ai-agent-auth-new-requirements | requirement |
 | Dynamic Task Coordination Requirements for AI Agents | draft-cui-ai-agent-task | requirement |
 *...and 5 more*
 ---
-## 9. Legal Liability Attribution Protocols
+## 6. Multi-Agent Conflict Resolution
 | | |
 |---|---|
 | **Severity** | HIGH |
-| **Category** | Policy/governance |
+| **Category** | A2A protocols |
-| **Drafts in category** | 115 |
+| **Drafts in category** | 157 |
-Missing technical protocols for creating audit trails that can determine legal liability when autonomous agents cause harm. Governance drafts exist but not technical accountability mechanisms.
+No protocol exists for resolving conflicts when multiple autonomous agents have competing objectives or try to access the same resources simultaneously. Current A2A protocols assume cooperative scenarios.
-**Evidence:** 115 governance drafts but legal technology gap for liability attribution
+**Evidence:** Multi-agent systems in production environments frequently deadlock or exhibit emergent adversarial behavior
 ### Related Drafts
 **Keyword matches** (drafts mentioning gap topic):
- [draft-madhavan-aipref-displaybasedpref](https://datatracker.ietf.org/doc/draft-madhavan-aipref-displaybasedpref/) (score 2.5) — A Vocabulary for Controlling Usage of Content Collected by Search and AI Crawler
+- [draft-li-dmsc-mcps-agw](https://datatracker.ietf.org/doc/draft-li-dmsc-mcps-agw/) (score 3.5) — Multi-agent Collaboration Protocol Suite based on Agent Gateway
- [draft-farzdusa-aipref-enduser](https://datatracker.ietf.org/doc/draft-farzdusa-aipref-enduser/) (score 3.8) — AI Preferences Signaling: End User Impact
+- [draft-li-dmsc-inf-architecture](https://datatracker.ietf.org/doc/draft-li-dmsc-inf-architecture/) (score 3.1) — Dynamic Multi-agent Secured Collaboration Infrastructure Architecture
- [draft-kotecha-agentic-dispute-protocol](https://datatracker.ietf.org/doc/draft-kotecha-agentic-dispute-protocol/) (score 3.6) — Agentic Dispute Protocol
+- [draft-yue-anima-agent-recovery-networks](https://datatracker.ietf.org/doc/draft-yue-anima-agent-recovery-networks/) (score 4.1) — Task-Oriented Multi-Agent Recovery Framework for High-Reliability in Converged M
- [draft-cui-dmsc-agent-cdi](https://datatracker.ietf.org/doc/draft-cui-dmsc-agent-cdi/) (score 3.0) — Cross-Domain Interoperability Framework for AI Agent Collaboration
+- [draft-chang-agent-context-interaction](https://datatracker.ietf.org/doc/draft-chang-agent-context-interaction/) (score 2.9) — Agent Context Interaction Optimizations
- [draft-ietf-aipref-vocab](https://datatracker.ietf.org/doc/draft-ietf-aipref-vocab/) (score 4.4) — A Vocabulary For Expressing AI Usage Preferences
+- [draft-fu-nmop-agent-communication-framework](https://datatracker.ietf.org/doc/draft-fu-nmop-agent-communication-framework/) (score 3.0) — Agent Communication Framework for Network AIOps
- [draft-aylward-aiga-1](https://datatracker.ietf.org/doc/draft-aylward-aiga-1/) (score 4.2) — AI Governance and Accountability Protocol (AIGA)
+- [draft-jurkovikj-httpapi-agentic-state](https://datatracker.ietf.org/doc/draft-jurkovikj-httpapi-agentic-state/) (score 3.9) — HTTP Profile for Synchronized Resource State (Agentic State Transfer)
-**Top-rated in Policy/governance** (115 drafts):
+**Top-rated in A2A protocols** (157 drafts):
- [draft-cowles-volt](https://datatracker.ietf.org/doc/draft-cowles-volt/) (4.8) — Defines tamper-evident execution trace format for AI agent workflows using hash chains and cryptogra
+- [draft-guy-bary-stamp-protocol](https://datatracker.ietf.org/doc/draft-guy-bary-stamp-protocol/) (4.6) — Defines STAMP protocol for cryptographic delegation and proof in AI agent systems. Provides task-bou
- [draft-aylward-daap-v2](https://datatracker.ietf.org/doc/draft-aylward-daap-v2/) (4.8) — Defines comprehensive protocol for AI agent accountability including authentication, monitoring, and
+- [draft-williams-netmod-lm-hierarchy-topology](https://datatracker.ietf.org/doc/draft-williams-netmod-lm-hierarchy-topology/) (4.6) — Defines YANG data model for hierarchical language model coordination across tiny, small, and large L
- [draft-goswami-agentic-jwt](https://datatracker.ietf.org/doc/draft-goswami-agentic-jwt/) (4.5) — Extends OAuth 2.0 with Agentic JWT to address authorization challenges in autonomous AI systems. Int
+- [draft-ietf-lake-edhoc](https://datatracker.ietf.org/doc/draft-ietf-lake-edhoc/) (4.6) — Specifies EDHOC, a compact authenticated Diffie-Hellman key exchange protocol for constrained enviro
- [draft-wang-cats-odsi](https://datatracker.ietf.org/doc/draft-wang-cats-odsi/) (4.5) — Specifies framework for decentralized LLM inference across untrusted participants with layer-aware e
+- [draft-chang-agent-token-efficient](https://datatracker.ietf.org/doc/draft-chang-agent-token-efficient/) (4.5) — Defines ADOL (Agentic Data Optimization Layer) to address token bloat in agent communication protoco
- [draft-birkholz-verifiable-agent-conversations](https://datatracker.ietf.org/doc/draft-birkholz-verifiable-agent-conversations/) (4.5) — Defines CDDL-based data format for verifiable agent conversation records using COSE signing. Support
+- [iso-23725-2024](https://datatracker.ietf.org/doc/iso-23725-2024/) (4.4) — Defines interoperability interfaces between fleet management and autonomous haulage systems in surfa
 ### Partially Addressing Ideas
@@ -412,85 +304,223 @@ No directly related technical ideas found in current drafts — this gap is enti
 ---
-## 10. Agent Memory and State Persistence Standards
+## 7. Inter-Standards-Body Protocol Bridging
 | | |
 |---|---|
-| **Severity** | MEDIUM |
+| **Severity** | HIGH |
-| **Category** | Data formats/interop |
+| **Category** | data formats/interop |
-| **Drafts in category** | 165 |
+| **Drafts in category** | 214 |
-No standardized formats or protocols exist for how agents should persist long-term memory, experience, and learned behaviors across system restarts or migrations. Each implementation creates proprietary solutions.
+Protocols developed by different standards bodies (IETF, ITU-T, ISO) cannot interoperate, creating silos where agents using ITU-T frameworks cannot communicate with those following IETF protocols.
-**Evidence:** 165 data format drafts focus on communication but not persistent state management
+**Evidence:** Enterprise environments need single agents to work across telecom networks (ITU-T), internet protocols (IETF), and industrial systems (ISO)
 ### Related Drafts
 **Keyword matches** (drafts mentioning gap topic):
 - [draft-cui-nmrg-auto-test](https://datatracker.ietf.org/doc/draft-cui-nmrg-auto-test/) (score 3.6) — Framework and Automation Levels for AI-Assisted Network Protocol Testing
 - [draft-cosmos-protocol-specification](https://datatracker.ietf.org/doc/draft-cosmos-protocol-specification/) (score 3.3) — The Cosmos Protocol Specification (Trust-Native Semantic Protocol)
 - [draft-tu-nmrg-blockchain-trusted-protocol](https://datatracker.ietf.org/doc/draft-tu-nmrg-blockchain-trusted-protocol/) (score 1.8) — A Blockchain Trusted Protocol for Intelligent Communication Network
 - [draft-yang-ioa-protocol](https://datatracker.ietf.org/doc/draft-yang-ioa-protocol/) (score 3.6) — Internet of Agents Protocol (IoA Protocol) for Heterogeneous Agent Collaboration
 - [draft-yang-dmsc-ioa-task-protocol](https://datatracker.ietf.org/doc/draft-yang-dmsc-ioa-task-protocol/) (score 3.0) — Internet of Agents Task Protocol (IoA Task Protocol) for Heterogeneous Agent Col
 - [draft-song-dmsc-problem-statement](https://datatracker.ietf.org/doc/draft-song-dmsc-problem-statement/) (score 3.0) — Problem Statement and Requirements for Dynamic Multi-agent Secured Collaboration
 **Top-rated in data formats/interop** (214 drafts):
 - [draft-cowles-volt](https://datatracker.ietf.org/doc/draft-cowles-volt/) (4.8) — Defines tamper-evident execution trace format for AI agent workflows using hash chains and cryptogra
 - [itu-t-y-3172](https://datatracker.ietf.org/doc/itu-t-y-3172/) (4.7) — Establishes comprehensive architectural framework for machine learning deployment in future networks
 - [iso-iec-22989-2022](https://datatracker.ietf.org/doc/iso-iec-22989-2022/) (4.7) — ISO/IEC standard defining core AI concepts and establishing standardized terminology across the fiel
 - [draft-williams-netmod-lm-hierarchy-topology](https://datatracker.ietf.org/doc/draft-williams-netmod-lm-hierarchy-topology/) (4.6) — Defines YANG data model for hierarchical language model coordination across tiny, small, and large L
 - [draft-ietf-lake-app-profiles](https://datatracker.ietf.org/doc/draft-ietf-lake-app-profiles/) (4.6) — Defines canonical CBOR representation for EDHOC application profiles and coordination mechanisms for
 ### Partially Addressing Ideas
 No directly related technical ideas found in current drafts — this gap is entirely unaddressed.
 ---
 ## 8. Agent Behavioral Audit Trails
 | | |
 |---|---|
 | **Severity** | HIGH |
 | **Category** | policy/governance |
 | **Drafts in category** | 214 |
 Missing standards for maintaining immutable logs of agent decisions and actions that can support forensic analysis and regulatory compliance. Current logging focuses on system events not decision rationale.
 **Evidence:** Financial and healthcare regulations require detailed audit trails, but AI systems cannot provide decision-level accountability
 ### Related Drafts
 **Keyword matches** (drafts mentioning gap topic):
 - [draft-zheng-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-agent-identity-management/) (score 3.7) — Agent Identity Managenment
 - [draft-li-dmsc-macp](https://datatracker.ietf.org/doc/draft-li-dmsc-macp/) (score 4.2) — Multi-agent Collaboration Protocol Suite
 - [draft-zheng-dispatch-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-dispatch-agent-identity-management/) (score 3.3) — Agent Identity Managenment
 - [draft-li-dmsc-macp](https://datatracker.ietf.org/doc/draft-li-dmsc-macp/) (score 4.2) — Multi-agent Collaboration Protocol Suite
 - [draft-fu-nmop-agent-communication-framework](https://datatracker.ietf.org/doc/draft-fu-nmop-agent-communication-framework/) (score 3.0) — Agent Communication Framework for Network AIOps
 - [draft-zyyhl-agent-networks-framework](https://datatracker.ietf.org/doc/draft-zyyhl-agent-networks-framework/) (score 3.6) — Framework for AI Agent Networks
- [draft-gaikwad-llm-benchmarking-terminology](https://datatracker.ietf.org/doc/draft-gaikwad-llm-benchmarking-terminology/) (score 2.7) — Benchmarking Terminology for Large Language Model Serving
+- [draft-kamimura-rats-behavioral-evidence](https://datatracker.ietf.org/doc/draft-kamimura-rats-behavioral-evidence/) (score 3.0) — On the Relationship Between Remote Attestation and Behavioral Evidence Recording
-**Top-rated in Data formats/interop** (165 drafts):
+**Top-rated in policy/governance** (214 drafts):
 - [draft-cowles-volt](https://datatracker.ietf.org/doc/draft-cowles-volt/) (4.8) — Defines tamper-evident execution trace format for AI agent workflows using hash chains and cryptogra
- [draft-williams-netmod-lm-hierarchy-topology](https://datatracker.ietf.org/doc/draft-williams-netmod-lm-hierarchy-topology/) (4.6) — Defines YANG data model for hierarchical language model coordination across tiny, small, and large L
+- [draft-aylward-daap-v2](https://datatracker.ietf.org/doc/draft-aylward-daap-v2/) (4.8) — Defines comprehensive protocol for AI agent accountability including authentication, monitoring, and
- [draft-ietf-lake-app-profiles](https://datatracker.ietf.org/doc/draft-ietf-lake-app-profiles/) (4.6) — Defines canonical CBOR representation for EDHOC application profiles and coordination mechanisms for
+- [itu-t-y-3172](https://datatracker.ietf.org/doc/itu-t-y-3172/) (4.7) — Establishes comprehensive architectural framework for machine learning deployment in future networks
- [draft-chang-agent-token-efficient](https://datatracker.ietf.org/doc/draft-chang-agent-token-efficient/) (4.5) — Defines ADOL (Agentic Data Optimization Layer) to address token bloat in agent communication protoco
+- [iso-iec-22989-2022](https://datatracker.ietf.org/doc/iso-iec-22989-2022/) (4.7) — ISO/IEC standard defining core AI concepts and establishing standardized terminology across the fiel
- [draft-birkholz-verifiable-agent-conversations](https://datatracker.ietf.org/doc/draft-birkholz-verifiable-agent-conversations/) (4.5) — Defines CDDL-based data format for verifiable agent conversation records using COSE signing. Support
+- [iso-iec-42001-2023](https://datatracker.ietf.org/doc/iso-iec-42001-2023/) (4.6) — ISO/IEC standard establishing comprehensive AI management system requirements covering governance, r
 ### Partially Addressing Ideas
-18 extracted ideas touch on this gap:
+9 extracted ideas touch on this gap:
 | Idea | Draft | Type |
 |------|-------|------|
 | Compliance-oriented agent memory model | draft-gaikwad-aps-profile | pattern |
-| Zero Trust Interoperability Framework | draft-liu-saag-zt-problem-statement | requirement |
+| Delegated Agent Authorization Protocol | draft-mishra-oauth-agent-grants | protocol |
-| Intelligent Agent Communication Gateway Architecture | draft-agent-gw | architecture |
+| Distributed AI Accountability Protocol | draft-aylward-daap-v2 | protocol |
-| Zero Trust Runtime Agent Architecture | draft-berlinai-vera | architecture |
+| Verifiable Agent Conversation Format | draft-birkholz-verifiable-agent-conversations | protocol |
-| Agentic Hypercall Protocol | draft-campbell-agentic-http | pattern |
+| Intent-Based Just-in-Time Authorization | draft-chen-agent-decoupled-authorization-model | architecture |
 | Dynamic Behavior-Based Authentication and Authorization Requirements | draft-chen-ai-agent-auth-new-requirements | requirement |
 | Agent Persistent State Profile | draft-gaikwad-aps-profile | architecture |
-| Agentic AI for Autonomous Network Management | draft-hong-nmrg-agenticai-ps | requirement |
+| Agent Interaction & Delegation Protocol | draft-vandoulas-aidp | protocol |
 | LISP-based geospatial intelligence network | draft-ietf-lisp-nexagon | protocol |
-*...and 10 more*
+*...and 1 more*
 ---
-## 11. Agent-to-Human Escalation Standards
+## 9. Agent Resource Consumption Limits
 | | |
 |---|---|
 | **Severity** | MEDIUM |
-| **Category** | Human-agent interaction |
+| **Category** | autonomous netops |
-| **Drafts in category** | 41 |
+| **Drafts in category** | 124 |
-While human-in-the-loop protocols exist, there's no standardized framework for when and how agents should escalate decisions to humans based on uncertainty, risk, or ethical considerations.
+Missing standards for how agents should self-regulate computational, network, and energy resource usage to prevent runaway consumption. Current traffic management focuses on traditional workloads, not autonomous agent behavior patterns.
-**Evidence:** Only 41 human-agent interaction drafts versus complex autonomous systems requiring escalation
+**Evidence:** Early agent deployments show unpredictable resource usage spikes that can destabilize infrastructure
 ### Related Drafts
 **Keyword matches** (drafts mentioning gap topic):
- [draft-williams-netmod-lm-hierarchy-topology](https://datatracker.ietf.org/doc/draft-williams-netmod-lm-hierarchy-topology/) (score 4.6) — Hierarchical Topology for Language Model Coordination
+- [draft-zheng-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-agent-identity-management/) (score 3.7) — Agent Identity Managenment
- [draft-ietf-websec-mime-sniff](https://datatracker.ietf.org/doc/draft-ietf-websec-mime-sniff/) (score 3.7) — Media Type Sniffing
+- [draft-zheng-dispatch-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-dispatch-agent-identity-management/) (score 3.3) — Agent Identity Managenment
- [draft-scrm-aiproto-usecases](https://datatracker.ietf.org/doc/draft-scrm-aiproto-usecases/) (score 4.1) — Agentic AI Use Cases
+- [draft-li-dmsc-macp](https://datatracker.ietf.org/doc/draft-li-dmsc-macp/) (score 4.2) — Multi-agent Collaboration Protocol Suite
- [draft-zeng-opsawg-llm-netconf-gap](https://datatracker.ietf.org/doc/draft-zeng-opsawg-llm-netconf-gap/) (score 3.9) — Gap Analysis of Network Configuration Protocols in LLM-Driven Intent-Based Netwo
+- [draft-fu-nmop-agent-communication-framework](https://datatracker.ietf.org/doc/draft-fu-nmop-agent-communication-framework/) (score 3.0) — Agent Communication Framework for Network AIOps
- [draft-jadoon-nmrg-agentic-ai-autonomous-networks](https://datatracker.ietf.org/doc/draft-jadoon-nmrg-agentic-ai-autonomous-networks/) (score 4.1) — Agentic AI Architectural Principles for Autonomous Computer Networks
+- [draft-zyyhl-agent-networks-framework](https://datatracker.ietf.org/doc/draft-zyyhl-agent-networks-framework/) (score 3.6) — Framework for AI Agent Networks
 - [draft-jia-oauth-scope-aggregation](https://datatracker.ietf.org/doc/draft-jia-oauth-scope-aggregation/) (score 3.5) — OAuth 2.0 Scope Aggregation for Multi-Step AI Agent Workflows
-**Top-rated in Human-agent interaction** (41 drafts):
+**Top-rated in autonomous netops** (124 drafts):
- [draft-drake-email-tpm-attestation](https://datatracker.ietf.org/doc/draft-drake-email-tpm-attestation/) (4.6) — Defines hardware attestation for email using TPM verification chains to prevent spam and provide Syb
+- [itu-t-y-3172](https://datatracker.ietf.org/doc/itu-t-y-3172/) (4.7) — Establishes comprehensive architectural framework for machine learning deployment in future networks
- [draft-ietf-aipref-vocab](https://datatracker.ietf.org/doc/draft-ietf-aipref-vocab/) (4.4) — Defines a standardized vocabulary for expressing preferences about how digital assets should be used
+- [iso-pas-8800-2024](https://datatracker.ietf.org/doc/iso-pas-8800-2024/) (4.5) — Addresses safety-related E/E systems using AI technology in series-production road vehicles, coverin
- [draft-dhir-http-agent-profile](https://datatracker.ietf.org/doc/draft-dhir-http-agent-profile/) (4.2) — Defines HTTP Agent Profile for authenticating agent traffic, separating human from agent traffic, an
+- [draft-cui-nmrg-llm-benchmark](https://datatracker.ietf.org/doc/draft-cui-nmrg-llm-benchmark/) (4.3) — Provides comprehensive evaluation framework for LLM-based network configuration agents. Includes emu
- [draft-song-tsvwg-camp](https://datatracker.ietf.org/doc/draft-song-tsvwg-camp/) (4.2) — Proposes CAMP, a multipath transport protocol for interactive multimodal LLM systems that maintains 
+- [iso-22733-1-2021](https://datatracker.ietf.org/doc/iso-22733-1-2021/) (4.3) — Specifies test methodology for evaluating autonomous emergency braking system performance in car-to-
- [draft-liu-agent-operation-authorization](https://datatracker.ietf.org/doc/draft-liu-agent-operation-authorization/) (4.1) — Specifies framework for verifiable delegation of actions from humans to AI agents using JWT tokens. 
+- [draft-wmz-nmrg-agent-ndt-arch](https://datatracker.ietf.org/doc/draft-wmz-nmrg-agent-ndt-arch/) (4.2) — Comprehensive architecture combining Network Digital Twin with Agentic AI for intent-based network o
 ### Partially Addressing Ideas
 12 extracted ideas touch on this gap:
 | Idea | Draft | Type |
 |------|-------|------|
 | SCIM 2.0 Extension for Agents and Agentic Applications | draft-abbey-scim-agent-extension | extension |
 | Context Distribution Optimization Framework | draft-chang-agent-context-interaction | mechanism |
 | Events Query Protocol | draft-gupta-httpapi-events-query | protocol |
 | Micro Agent Communication Protocol (µACP) | draft-mallick-muacp | protocol |
 | MOQT Binding for A2A and MCP Protocols | draft-nandakumar-ai-agent-moq-transport | extension |
 | AI Agent Protocol Requirements | draft-rosenberg-ai-protocols | requirement |
 | SCIM 2.0 Agent Extension | draft-scim-agent-extension | extension |
 | Authorized Connection Policy Framework | draft-steckbeck-ua-conn-sec | mechanism |
 *...and 4 more*
 ---
 ## 10. Agent Training Data Provenance Tracking
 | | |
 |---|---|
 | **Severity** | MEDIUM |
 | **Category** | data formats/interop |
 | **Drafts in category** | 214 |
 Missing standards for tracking the lineage and provenance of training data as it flows between agents and gets updated through federated learning or agent interactions.
 **Evidence:** Data protection regulations require knowing data sources, but current AI systems cannot trace training data origins
 ### Related Drafts
 **Keyword matches** (drafts mentioning gap topic):
 - [draft-zheng-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-agent-identity-management/) (score 3.7) — Agent Identity Managenment
 - [draft-zheng-dispatch-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-dispatch-agent-identity-management/) (score 3.3) — Agent Identity Managenment
 - [draft-li-dmsc-macp](https://datatracker.ietf.org/doc/draft-li-dmsc-macp/) (score 4.2) — Multi-agent Collaboration Protocol Suite
 - [draft-fu-nmop-agent-communication-framework](https://datatracker.ietf.org/doc/draft-fu-nmop-agent-communication-framework/) (score 3.0) — Agent Communication Framework for Network AIOps
 - [draft-zyyhl-agent-networks-framework](https://datatracker.ietf.org/doc/draft-zyyhl-agent-networks-framework/) (score 3.6) — Framework for AI Agent Networks
 - [draft-akhavain-moussa-ai-network](https://datatracker.ietf.org/doc/draft-akhavain-moussa-ai-network/) (score 3.9) — AI Network for Training, Inference, and Agentic Interactions
 **Top-rated in data formats/interop** (214 drafts):
 - [draft-cowles-volt](https://datatracker.ietf.org/doc/draft-cowles-volt/) (4.8) — Defines tamper-evident execution trace format for AI agent workflows using hash chains and cryptogra
 - [itu-t-y-3172](https://datatracker.ietf.org/doc/itu-t-y-3172/) (4.7) — Establishes comprehensive architectural framework for machine learning deployment in future networks
 - [iso-iec-22989-2022](https://datatracker.ietf.org/doc/iso-iec-22989-2022/) (4.7) — ISO/IEC standard defining core AI concepts and establishing standardized terminology across the fiel
 - [draft-williams-netmod-lm-hierarchy-topology](https://datatracker.ietf.org/doc/draft-williams-netmod-lm-hierarchy-topology/) (4.6) — Defines YANG data model for hierarchical language model coordination across tiny, small, and large L
 - [draft-ietf-lake-app-profiles](https://datatracker.ietf.org/doc/draft-ietf-lake-app-profiles/) (4.6) — Defines canonical CBOR representation for EDHOC application profiles and coordination mechanisms for
 ### Partially Addressing Ideas
 30 extracted ideas touch on this gap:
 | Idea | Draft | Type |
 |------|-------|------|
 | EAT Profile for AI Agent Attestation | draft-messous-eat-ai | extension |
 | Warrant Certificate Authority (WCA) | draft-bondar-wca | architecture |
 | Blockchain-Anchored Integrity for AI Provenance | draft-reilly-sentinel-protocol | mechanism |
 | AI Traffic Characterization Framework | draft-aft-ai-traffic | requirement |
 | AI Traffic Characterization Framework | draft-ai-traffic | requirement |
 | Network Architecture for AI Training and Inference | draft-akhavain-moussa-ai-network | architecture |
 | Verifiable Agent Conversation Format | draft-birkholz-verifiable-agent-conversations | protocol |
 | Agentic Data Optimization Layer (ADOL) | draft-chang-agent-token-efficient | protocol |
 *...and 22 more*
 ---
 ## 11. Agent-Generated Content Attribution
 | | |
 |---|---|
 | **Severity** | MEDIUM |
 | **Category** | data formats/interop |
 | **Drafts in category** | 214 |
 Missing technical standards for embedding cryptographic attribution in content created by agents, enabling detection of AI-generated text, code, or decisions. Current synthetic content guidance lacks implementation details.
 **Evidence:** Need to distinguish agent-generated content from human-generated for legal, security, and quality assurance purposes
 ### Related Drafts
 **Keyword matches** (drafts mentioning gap topic):
 - [draft-romanchuk-normative-admissibility](https://datatracker.ietf.org/doc/draft-romanchuk-normative-admissibility/) (score 3.4) — Normative Admissibility Framework for Agent Speech Acts
 - [draft-li-semantic-routing-architecture](https://datatracker.ietf.org/doc/draft-li-semantic-routing-architecture/) (score 3.6) — Semantic Routing Architecture for AI Agents Communication
 - [draft-cui-nmrg-llm-nm](https://datatracker.ietf.org/doc/draft-cui-nmrg-llm-nm/) (score 4.1) — A Framework for LLM Agent-Assisted Network Management with Human-in-the-Loop
 - [draft-mpsb-agntcy-messaging](https://datatracker.ietf.org/doc/draft-mpsb-agntcy-messaging/) (score 2.6) — An Overview of Messaging Systems and Their Applicability to Agentic AI
 - [draft-bondar-wca](https://datatracker.ietf.org/doc/draft-bondar-wca/) (score 3.9) — Warrant Certificate Authorities (WCA): Auditable Data Provenance for AI-Agent To
 - [draft-it-aipref-attachment](https://datatracker.ietf.org/doc/draft-it-aipref-attachment/) (score 3.2) — Indicating Preferences Regarding Content Usage
 **Top-rated in data formats/interop** (214 drafts):
 - [draft-cowles-volt](https://datatracker.ietf.org/doc/draft-cowles-volt/) (4.8) — Defines tamper-evident execution trace format for AI agent workflows using hash chains and cryptogra
 - [itu-t-y-3172](https://datatracker.ietf.org/doc/itu-t-y-3172/) (4.7) — Establishes comprehensive architectural framework for machine learning deployment in future networks
 - [iso-iec-22989-2022](https://datatracker.ietf.org/doc/iso-iec-22989-2022/) (4.7) — ISO/IEC standard defining core AI concepts and establishing standardized terminology across the fiel
 - [draft-williams-netmod-lm-hierarchy-topology](https://datatracker.ietf.org/doc/draft-williams-netmod-lm-hierarchy-topology/) (4.6) — Defines YANG data model for hierarchical language model coordination across tiny, small, and large L
 - [draft-ietf-lake-app-profiles](https://datatracker.ietf.org/doc/draft-ietf-lake-app-profiles/) (4.6) — Defines canonical CBOR representation for EDHOC application profiles and coordination mechanisms for
 ### Partially Addressing Ideas
@@ -498,57 +528,19 @@ No directly related technical ideas found in current drafts — this gap is enti
 ---
 ## 12. Federated Agent Learning Privacy
 | | |
 |---|---|
 | **Severity** | MEDIUM |
 | **Category** | new |
 | **Drafts in category** | 0 |
 Federated AI operations models exist but lack privacy-preserving protocols for agents learning from shared experiences without exposing sensitive data from individual deployments.
 **Evidence:** Federated models mentioned but privacy-preserving learning protocols absent
 ### Related Drafts
 **Keyword matches** (drafts mentioning gap topic):
 - [draft-kale-agntcy-federated-privacy](https://datatracker.ietf.org/doc/draft-kale-agntcy-federated-privacy/) (score 3.2) — Privacy-Preserving Federated Learning Architecture for Multi-Tenant AI Agent Sys
 - [draft-cui-dmsc-agent-cdi](https://datatracker.ietf.org/doc/draft-cui-dmsc-agent-cdi/) (score 3.0) — Cross-Domain Interoperability Framework for AI Agent Collaboration
 - [draft-ai-traffic](https://datatracker.ietf.org/doc/draft-ai-traffic/) (score 2.5) — Handling inter-DC/Edge AI-related network traffic: Problem statement
 - [draft-aft-ai-traffic](https://datatracker.ietf.org/doc/draft-aft-ai-traffic/) (score 3.1) — Handling inter-DC/Edge AI-related network traffic: Problem statement
 - [draft-aylward-aiga-1](https://datatracker.ietf.org/doc/draft-aylward-aiga-1/) (score 4.2) — AI Governance and Accountability Protocol (AIGA)
 - [draft-zheng-agent-identity-management](https://datatracker.ietf.org/doc/draft-zheng-agent-identity-management/) (score 3.7) — Agent Identity Managenment
 ### Partially Addressing Ideas
 5 extracted ideas touch on this gap:
 | Idea | Draft | Type |
 |------|-------|------|
 | Privacy-Preserving Federated Learning for Multi-Tenant AI Agents | draft-kale-agntcy-federated-privacy | architecture |
 | Cross-Domain Agent Interoperability Framework | draft-cui-dmsc-agent-cdi | architecture |
 | HTTP Agent Profile (HAP) | draft-dhir-http-agent-profile | protocol |
 | AI Network Security Agent | draft-yuan-rtgwg-security-agent-usecase | architecture |
 | AI Network Traffic Optimization Agent | draft-yuan-rtgwg-traffic-agent-usecase | architecture |
 ---
 ## Cross-Cutting Analysis
 ### Gaps by Category
 | Category | Drafts | Gaps | Gap Topics |
 |----------|-------:|-----:|------------|
-| a2a protocols | 150 | 2 | Multi-Agent Consensus Under Byzantine Conditions; Cross-Protocol Agent Migration |
+| a2a protocols | 157 | 1 | Multi-Agent Conflict Resolution |
-| agent discovery/reg | 87 | 1 | Agent Capability Advertisement Verification |
+| agent identity/auth | 160 | 1 | Cross-Domain Agent Identity Portability |
-| agent identity/auth | 145 | 1 | Cross-Domain Agent Communication Security |
+| ai safety/alignment | 112 | 2 | Agent Capability Degradation Detection; Emergency Agent Override Protocols |
-| ai safety/alignment | 46 | 2 | Real-time Agent Behavior Verification; Emergency Agent Shutdown Coordination |
+| autonomous netops | 124 | 1 | Agent Resource Consumption Limits |
-| data formats/interop | 165 | 1 | Agent Memory and State Persistence Standards |
+| data formats/interop | 214 | 3 | Inter-Standards-Body Protocol Bridging; Agent Training Data Provenance Tracking; Agent-Generated Content Attribution |
-| human-agent interaction | 41 | 1 | Agent-to-Human Escalation Standards |
+| human-agent interaction | 57 | 1 | Real-Time Agent Behavior Explanation |
-| new | 0 | 3 | Agent Resource Accounting and Billing; Agent Performance Degradation Detection; Federated Agent Learning Privacy |
+| policy/governance | 214 | 2 | Agent Legal Liability Framework; Agent Behavioral Audit Trails |
 | policy/governance | 115 | 1 | Legal Liability Attribution Protocols |
 ## Recommendations
--- a/scripts/fetch-all-sources.py
+++ b/scripts/fetch-all-sources.py
@@ -15,7 +15,7 @@ cfg = Config.load()
 db = Database(cfg)
 # Only fetch from new sources (IETF and W3C already done recently)
-sources_to_fetch = ["etsi", "itu", "iso"]
+sources_to_fetch = ["etsi", "itu", "iso", "nist"]
 total_new = 0
 for source_name in sources_to_fetch:
--- a/src/ietf_analyzer/analyzer.py
+++ b/src/ietf_analyzer/analyzer.py
@@ -160,29 +160,34 @@ Return ONLY a JSON object like {{"draft-name":[...], ...}}, no fences."""
 # independent gap analyses and intersect results, have domain experts validate.
 # ============================================================================
 GAP_ANALYSIS_PROMPT = """\
-You are analyzing the landscape of {total} IETF Internet-Drafts related to AI agents and autonomous systems.
+You are analyzing the landscape of {total} documents related to AI agents and autonomous systems from multiple standards organizations.
-## Categories and Draft Counts
+## IETF Drafts — Categories and Draft Counts
 {category_summary}
-## Most Common Technical Ideas
+## Most Common Technical Ideas (from IETF drafts)
 {top_ideas}
 ## Known Overlap Clusters (groups of highly similar drafts)
 {overlap_summary}
-Identify 8-15 GAPS — areas, problems, or technical challenges NOT adequately addressed by existing drafts.
+## Other Standards Bodies
 {other_sources_summary}
 Identify 8-15 GAPS — areas, problems, or technical challenges NOT adequately addressed by existing drafts and standards.
 Return a JSON array:
-[{{"topic":"short topic name","description":"2-3 sentence description","category":"closest category or new","severity":"critical|high|medium|low","evidence":"what suggests this gap matters"}}]
+[{{"topic":"short topic name","description":"2-3 sentence description","category":"closest category or new","severity":"critical|high|medium|low","evidence":"what suggests this gap matters","addressed_by":"which existing standards (if any) partially address this, from any source"}}]
 Focus on:
-1. Problems mentioned but not solved
+1. Problems mentioned but not solved — even across organizations
-2. Missing infrastructure pieces
+2. Missing infrastructure pieces (no standard from ANY body covers it)
 3. Security/privacy/safety issues not addressed
-4. Interoperability gaps between competing proposals
+4. Interoperability gaps between competing proposals or between standards bodies
 5. Real-world deployment concerns ignored
 6. Cross-organization coordination gaps (e.g., IETF protocol needs ISO governance framework)
 Consider what NIST, ISO, ETSI, ITU-T, and W3C already cover vs what remains missing.
 JSON array only, no fences."""
 SCORE_NOVELTY_PROMPT = """\
@@ -638,11 +643,31 @@ class Analyzer:
        for c, n in sorted(cat_counts.items(), key=lambda x: x[1], reverse=True)[:5]:
            overlap_summary += f"- {c} ({n} drafts, high internal overlap)\n"
        # Build summary of non-IETF sources
        other_rows = self.db.conn.execute(
            "SELECT source, name, title, abstract FROM drafts WHERE source != 'ietf' ORDER BY source, name"
        ).fetchall()
        source_groups: dict[str, list] = {}
        for r in other_rows:
            src = r["source"].upper()
            source_groups.setdefault(src, []).append(r)
        other_lines = []
        for src, docs in sorted(source_groups.items()):
            other_lines.append(f"\n### {src} ({len(docs)} documents)")
            for d in docs[:30]:  # cap per source to fit context
                abstract = (d["abstract"] or "")[:150]
                other_lines.append(f"- **{d['title'][:100]}**: {abstract}")
            if len(docs) > 30:
                other_lines.append(f"  ... and {len(docs) - 30} more")
        other_sources_summary = "\n".join(other_lines) if other_lines else "(No other sources available)"
        prompt = GAP_ANALYSIS_PROMPT.format(
            total=total,
            category_summary=category_summary,
            top_ideas=top_ideas,
            overlap_summary=overlap_summary,
            other_sources_summary=other_sources_summary,
        )
        phash = _prompt_hash(prompt)
--- a/src/ietf_analyzer/cli.py
+++ b/src/ietf_analyzer/cli.py
@@ -3081,3 +3081,255 @@ def export(export_type: str, fmt: str, output_file: str | None):
    finally:
        db.close()
 # ── auto ─────────────────────────────────────────────────────────────────────
@main.command("auto")
@click.option("--cost-limit", default=2.0, help="Auto-approve operations under this USD amount (default: $2)")
@click.option("--yes", "-y", is_flag=True, help="Skip all confirmation prompts")
@click.option("--dry-run", is_flag=True, help="Show what would be done without doing it")
@click.option("--source", "-s", default=None, help="Limit to specific source (ietf,w3c,etsi,iso,itu)")
 def auto(cost_limit: float, yes: bool, dry_run: bool, source: str | None):
    """Auto-heal: fetch, analyze, embed, extract ideas, and update gaps.
    Automatically processes all unrated, unembedded, and idea-less drafts
    across all sources. Uses cheap models (Haiku) for bulk operations.
    Operations estimated above --cost-limit require confirmation.
    Examples:
        ietf auto                # run full pipeline, auto-approve under $2
        ietf auto --dry-run      # show plan without executing
        ietf auto -s iso         # only process ISO drafts
        ietf auto --cost-limit 5 # raise approval threshold to $5
        ietf auto -y             # skip all prompts (for cron)
    """
    cfg = Config.load()
    db = Database(cfg)
    try:
        _auto_heal(cfg, db, cost_limit=cost_limit, yes=yes, dry_run=dry_run, source_filter=source)
    finally:
        db.close()
 def _estimate_cost(n_drafts: int, operation: str) -> float:
    """Estimate USD cost for an operation. Conservative estimates."""
    # Haiku: ~$0.25/M input, ~$1.25/M output
    # Sonnet: ~$3/M input, ~$15/M output
    # Average draft abstract: ~500 tokens input, ~200 tokens output
    costs = {
        "analyze_cheap": n_drafts * 0.0005,   # ~$0.50 per 1000 drafts (Haiku)
        "analyze_quality": n_drafts * 0.005,   # ~$5.00 per 1000 drafts (Sonnet)
        "ideas_cheap": n_drafts * 0.001,       # ~$1.00 per 1000 drafts (Haiku batch)
        "ideas_quality": n_drafts * 0.008,     # ~$8.00 per 1000 drafts (Sonnet)
        "gaps": 0.05,                          # single Claude call
        "embed": 0.0,                          # Ollama is free/local
        "authors": 0.0,                        # Datatracker API is free
        "fetch": 0.0,                          # Datatracker API is free
    }
    return costs.get(operation, 0.0)
 def _auto_heal(cfg, db, cost_limit: float, yes: bool, dry_run: bool, source_filter: str | None):
    """Run the full auto-heal pipeline."""
    import time as _time
    from rich.panel import Panel
    steps: list[dict] = []
    total_cost = 0.0
    # ── Step 1: Fetch new drafts from all sources ──
    sources = [source_filter] if source_filter else cfg.observatory_sources
    steps.append({
        "name": f"Fetch new drafts from {', '.join(sources)}",
        "sources": sources,
        "cost": 0.0,
        "action": "fetch",
    })
    # ── Step 2: Analyze unrated drafts ──
    unrated = db.unrated_drafts(limit=10000)
    if source_filter:
        unrated = [d for d in unrated if (d.source or "ietf") == source_filter]
    n_unrated = len(unrated)
    analyze_cost = _estimate_cost(n_unrated, "analyze_cheap")
    steps.append({
        "name": f"Analyze {n_unrated} unrated drafts (Haiku)",
        "count": n_unrated,
        "cost": analyze_cost,
        "action": "analyze",
    })
    total_cost += analyze_cost
    # ── Step 3: Fetch authors ──
    missing_authors = db.conn.execute(
        "SELECT COUNT(*) FROM drafts WHERE name NOT IN (SELECT DISTINCT draft_name FROM draft_authors)"
    ).fetchone()[0]
    steps.append({
        "name": f"Fetch authors for {missing_authors} drafts",
        "count": missing_authors,
        "cost": 0.0,
        "action": "authors",
    })
    # ── Step 4: Embed missing drafts ──
    missing_embed = db.drafts_without_embeddings(limit=10000)
    if source_filter:
        source_names = {row[0] for row in db.conn.execute(
            "SELECT name FROM drafts WHERE source = ?", (source_filter,)
        ).fetchall()}
        missing_embed = [n for n in missing_embed if n in source_names]
    n_embed = len(missing_embed)
    steps.append({
        "name": f"Embed {n_embed} drafts (Ollama, free)",
        "count": n_embed,
        "cost": 0.0,
        "action": "embed",
    })
    # ── Step 5: Extract ideas ──
    missing_ideas = db.drafts_without_ideas(limit=10000)
    if source_filter:
        if not source_names:
            source_names = {row[0] for row in db.conn.execute(
                "SELECT name FROM drafts WHERE source = ?", (source_filter,)
            ).fetchall()}
        missing_ideas = [n for n in missing_ideas if n in source_names]
    n_ideas = len(missing_ideas)
    ideas_cost = _estimate_cost(n_ideas, "ideas_cheap")
    steps.append({
        "name": f"Extract ideas from {n_ideas} drafts (Haiku)",
        "count": n_ideas,
        "cost": ideas_cost,
        "action": "ideas",
    })
    total_cost += ideas_cost
    # ── Step 6: Refresh gaps ──
    gap_cost = _estimate_cost(0, "gaps")
    steps.append({
        "name": "Refresh gap analysis",
        "cost": gap_cost,
        "action": "gaps",
    })
    total_cost += gap_cost
    # ── Show plan ──
    plan_lines = []
    for s in steps:
        count = s.get("count", 1)
        if count == 0:
            plan_lines.append(f"  [dim]SKIP[/]  {s['name']}")
        else:
            cost_str = f"  [yellow]~${s['cost']:.2f}[/]" if s["cost"] > 0 else ""
            plan_lines.append(f"  [green]RUN[/]   {s['name']}{cost_str}")
    auto_approved = total_cost <= cost_limit
    plan_lines.append(f"\n  [bold]Estimated total cost: ${total_cost:.2f}[/]")
    if auto_approved:
        plan_lines.append(f"  [green]Auto-approved (under ${cost_limit:.2f} limit)[/]")
    else:
        plan_lines.append(f"  [yellow]Requires approval (over ${cost_limit:.2f} limit)[/]")
    console.print(Panel("\n".join(plan_lines), title="Auto-Heal Plan"))
    if dry_run:
        console.print("[bold yellow]DRY RUN[/] — no changes made.")
        return
    # ── Approval ──
    if not auto_approved and not yes:
        if not click.confirm(f"Estimated cost ${total_cost:.2f} exceeds ${cost_limit:.2f} limit. Proceed?"):
            console.print("[yellow]Aborted.[/]")
            return
    # ── Execute ──
    start = _time.time()
    for step in steps:
        action = step["action"]
        count = step.get("count", 0)
        if action == "fetch":
            console.print(f"\n[bold cyan]>>> Fetching from {step['sources']}...[/]")
            from .sources import get_fetcher
            from .observatory import _doc_to_draft
            for src_name in step["sources"]:
                try:
                    fetcher = get_fetcher(src_name, cfg)
                    before = db.count_drafts()
                    results = fetcher.search(keywords=cfg.search_keywords)
                    for doc in results:
                        db.upsert_draft(_doc_to_draft(doc))
                    after = db.count_drafts()
                    new = after - before
                    console.print(f"  [{src_name}] +{new} new drafts")
                    fetcher.close()
                except Exception as e:
                    console.print(f"  [{src_name}] [red]Error: {e}[/]")
        elif action == "analyze" and count > 0:
            console.print(f"\n[bold cyan]>>> Analyzing {count} drafts (Haiku)...[/]")
            from .analyzer import Analyzer
            analyzer = Analyzer(cfg, db)
            orig_model = cfg.claude_model
            cfg.claude_model = cfg.claude_model_cheap
            try:
                done = analyzer.rate_all_unrated(limit=count)
                console.print(f"  Analyzed [bold green]{done}[/] drafts")
            finally:
                cfg.claude_model = orig_model
        elif action == "authors" and count > 0:
            console.print(f"\n[bold cyan]>>> Fetching authors for {count} drafts...[/]")
            from .authors import AuthorNetwork
            author_net = AuthorNetwork(cfg, db)
            done = author_net.fetch_all_authors()
            console.print(f"  Fetched authors for [bold green]{done}[/] drafts")
        elif action == "embed" and count > 0:
            console.print(f"\n[bold cyan]>>> Embedding {count} drafts (Ollama)...[/]")
            from .embeddings import Embedder
            with Embedder(cfg, db) as embedder:
                done = embedder.embed_all_missing()
                console.print(f"  Embedded [bold green]{done}[/] drafts")
        elif action == "ideas" and count > 0:
            console.print(f"\n[bold cyan]>>> Extracting ideas from {count} drafts (Haiku)...[/]")
            from .analyzer import Analyzer
            analyzer = Analyzer(cfg, db)
            done = analyzer.extract_all_ideas(limit=count, batch_size=5, cheap=True)
            console.print(f"  Extracted ideas from [bold green]{done}[/] drafts")
        elif action == "gaps":
            console.print(f"\n[bold cyan]>>> Refreshing gap analysis...[/]")
            from .analyzer import Analyzer
            analyzer = Analyzer(cfg, db)
            gaps = analyzer.gap_analysis()
            if gaps:
                console.print(f"  Found [bold green]{len(gaps)}[/] gaps")
    elapsed = _time.time() - start
    console.print(f"\n[bold green]Auto-heal complete![/] ({elapsed:.1f}s, ~${total_cost:.2f})")
    # Show final counts
    total = db.count_drafts()
    rated = db.conn.execute("SELECT COUNT(*) FROM ratings").fetchone()[0]
    embedded = db.conn.execute("SELECT COUNT(*) FROM embeddings").fetchone()[0]
    idea_count = db.conn.execute("SELECT COUNT(*) FROM ideas").fetchone()[0]
    gap_count = db.conn.execute("SELECT COUNT(*) FROM gaps").fetchone()[0]
    console.print(f"  Drafts: {total} | Rated: {rated} | Embedded: {embedded} | Ideas: {idea_count} | Gaps: {gap_count}")
    by_source = db.conn.execute(
        "SELECT source, COUNT(*) FROM drafts GROUP BY source ORDER BY COUNT(*) DESC"
    ).fetchall()
    source_str = " | ".join(f"{s}: {c}" for s, c in by_source)
    console.print(f"  Sources: {source_str}")
--- a/src/ietf_analyzer/config.py
+++ b/src/ietf_analyzer/config.py
@@ -52,7 +52,7 @@ class Config:
    # Observatory — add "w3c" to enable W3C spec tracking:
    #   ietf observatory update --source w3c    (one-off)
    #   or set observatory_sources to ["ietf", "w3c"] in config.json
-    observatory_sources: list[str] = field(default_factory=lambda: ["ietf", "w3c", "etsi", "itu", "iso"])
+    observatory_sources: list[str] = field(default_factory=lambda: ["ietf", "w3c", "etsi", "itu", "iso", "nist"])
    dashboard_dir: str = str(DEFAULT_DATA_DIR.parent / "docs")
    w3c_groups: list[str] = field(default_factory=lambda: [
        "webmachinelearning", "wot", "credentials", "did", "vc"
--- a/src/ietf_analyzer/sources/init.py
+++ b/src/ietf_analyzer/sources/init.py
@@ -5,6 +5,7 @@ from .etsi import ETSIFetcher
 from .ietf import IETFFetcher
 from .iso import ISOFetcher
 from .itu import ITUFetcher
 from .nist import NISTFetcher
 from .w3c import W3CFetcher
 FETCHERS = {
@@ -13,6 +14,7 @@ FETCHERS = {
    "etsi": ETSIFetcher,
    "itu": ITUFetcher,
    "iso": ISOFetcher,
    "nist": NISTFetcher,
 }
--- a/src/ietf_analyzer/sources/iso.py
+++ b/src/ietf_analyzer/sources/iso.py
@@ -151,10 +151,11 @@ class ISOFetcher:
                    continue
                name = _iso_id_to_name(ref)
                abstract = scope[:2000] if scope else f"ISO/IEC standard: {title}. Committee: {committee}."
                docs.append(SourceDocument(
                    name=name,
                    title=f"{ref}: {title}",
-                    abstract=f"ISO/IEC standard: {title}. Committee: {committee}. Status: {status}.",
+                    abstract=abstract,
                    source="iso",
                    source_id=ref,
                    source_url=f"https://www.iso.org/standard/{ref.split(':')[0].replace('/', '%2F').replace(' ', '%20')}.html",
--- a/src/webui/data.py
+++ b/src/webui/data.py
@@ -892,11 +892,13 @@ def _compute_author_network_full(db: Database) -> AuthorNetwork:
        if len(component) >= 2:
            org_mix: dict[str, int] = Counter()
            member_orgs: dict[str, str] = {}
            cluster_drafts: dict[str, str] = {}  # name -> title
            for m in component:
                org = author_info.get(m, {}).get("org", "")
                if org:
                    org_mix[org] += 1
                    member_orgs[m] = org
                for dn in author_info.get(m, {}).get("drafts", []):
                    if dn not in cluster_drafts:
                        d = _all_drafts_map.get(dn)
@@ -904,9 +906,10 @@ def _compute_author_network_full(db: Database) -> AuthorNetwork:
            clusters.append({
                "id": len(clusters),
                "members": component,
                "member_orgs": member_orgs,
                "org_mix": dict(org_mix.most_common()),
                "size": len(component),
-                "drafts": [{"name": n, "title": t} for n, t in list(cluster_drafts.items())[:15]],
+                "drafts": [{"name": n, "title": t} for n, t in list(cluster_drafts.items())],
                "draft_count": len(cluster_drafts),
            })
@@ -1062,11 +1065,78 @@ def _compute_idea_clusters(db: Database) -> dict:
    except Exception:
        pass
    # --- Cross-cluster links ---
    # Find pairs of clusters whose ideas are semantically related
    # Use centroid similarity + best idea-pair links
    links = []
    if len(clusters) >= 2:
        # Build cluster centroids from normalized embeddings
        cluster_centroids = {}
        cluster_member_indices: dict[int, list[int]] = defaultdict(list)
        for idx, iid in enumerate(idea_ids):
            cid = iid_to_new.get(iid, int(labels[idx]))
            cluster_member_indices[cid].append(idx)
        for cid, indices in cluster_member_indices.items():
            if indices:
                centroid = matrix_norm[indices].mean(axis=0)
                norm = np.linalg.norm(centroid)
                if norm > 0:
                    cluster_centroids[cid] = centroid / norm
        # Compute pairwise centroid similarity for all cluster pairs
        cids_sorted = sorted(cluster_centroids.keys())
        for ci_idx, ci in enumerate(cids_sorted):
            for cj in cids_sorted[ci_idx + 1:]:
                sim = float(np.dot(cluster_centroids[ci], cluster_centroids[cj]))
                if sim < 0.45:
                    continue
                # Find the best idea pair across these two clusters
                best_sim = 0.0
                best_pair = (None, None)
                # Sample up to 20 ideas per cluster to keep it fast
                ci_members = cluster_member_indices[ci][:20]
                cj_members = cluster_member_indices[cj][:20]
                for mi in ci_members:
                    for mj in cj_members:
                        pair_sim = float(np.dot(matrix_norm[mi], matrix_norm[mj]))
                        if pair_sim > best_sim:
                            best_sim = pair_sim
                            best_pair = (idea_ids[mi], idea_ids[mj])
                if best_sim < 0.5:
                    continue
                # Get theme names
                ci_theme = next((c["theme"] for c in clusters if c["id"] == ci), f"Cluster {ci}")
                cj_theme = next((c["theme"] for c in clusters if c["id"] == cj), f"Cluster {cj}")
                idea_a = idea_map.get(best_pair[0], {})
                idea_b = idea_map.get(best_pair[1], {})
                links.append({
                    "source": ci,
                    "target": cj,
                    "source_theme": ci_theme,
                    "target_theme": cj_theme,
                    "similarity": round(sim, 3),
                    "best_pair_sim": round(best_sim, 3),
                    "idea_a": idea_a.get("title", ""),
                    "idea_a_draft": idea_a.get("draft_name", ""),
                    "idea_b": idea_b.get("title", ""),
                    "idea_b_draft": idea_b.get("draft_name", ""),
                })
        links.sort(key=lambda l: l["best_pair_sim"], reverse=True)
        links = links[:50]  # cap at top 50 links
    total = len(idea_ids)
    clustered = sum(c["size"] for c in clusters)
    return {
        "clusters": clusters,
        "scatter": scatter,
        "links": links,
        "stats": {"total": total, "clustered": clustered, "num_clusters": len(clusters)},
        "empty": False,
    }
--- a/src/webui/templates/authors.html
+++ b/src/webui/templates/authors.html
@@ -116,34 +116,72 @@
    <p class="text-xs text-slate-500 mb-4">Clusters are formed by connected-component analysis of the co-authorship graph: authors who share 2+ drafts are linked, and all authors reachable through such links form a cluster. This reveals research teams and institutional collaboration patterns — a cluster of 20 authors from 3 organizations means those groups actively co-author across org boundaries. Click a cluster to highlight it in the graph above.</p>
    <div class="grid grid-cols-1 md:grid-cols-2 lg:grid-cols-3 gap-4" id="clusterGrid">
        {% for c in network.clusters[:12] %}
-        <div class="cluster-card bg-slate-800/50 rounded-lg border border-slate-700/50 p-4 cursor-pointer" data-cluster-id="{{ c.id }}" onclick="highlightCluster({{ c.id }})">
+        <div class="cluster-card bg-slate-800/50 rounded-lg border border-slate-700/50 p-4 cursor-pointer" data-cluster-id="{{ c.id }}">
-            <div class="flex items-center justify-between mb-2">
+            <!-- Header — click to highlight in graph -->
            <div class="flex items-center justify-between mb-2" onclick="highlightCluster({{ c.id }})">
                <span class="text-sm font-semibold text-white">Cluster #{{ c.id + 1 }}</span>
-                <div class="flex gap-1.5">
+                <div class="flex gap-1.5 items-center">
                    <span class="text-xs px-2 py-0.5 rounded-full bg-blue-500/20 text-blue-400">{{ c.size }} authors</span>
                    <span class="text-xs px-2 py-0.5 rounded-full bg-emerald-500/20 text-emerald-400">{{ c.draft_count }} drafts</span>
                    <svg class="w-4 h-4 text-slate-500 transition-transform cluster-chevron" fill="none" stroke="currentColor" viewBox="0 0 24 24"><path stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M19 9l-7 7-7-7"/></svg>
                </div>
            </div>
            <!-- Org mix -->
            <div class="flex flex-wrap gap-1 mb-2">
                {% for org, count in c.org_mix.items() %}
                <span class="text-xs px-2 py-0.5 rounded-full bg-slate-700 text-slate-300">{{ org }} ({{ count }})</span>
                {% endfor %}
            </div>
            <!-- Preview: first 3 members -->
            <div class="text-xs text-slate-500 mb-2 truncate" title="{{ c.members | join(', ') }}">
-                {{ c.members[:5] | join(', ') }}{% if c.members | length > 5 %} +{{ c.members | length - 5 }} more{% endif %}
+                {{ c.members[:3] | join(', ') }}{% if c.members | length > 3 %} +{{ c.members | length - 3 }} more{% endif %}
            </div>
            <!-- Preview: first 3 drafts -->
            {% if c.drafts %}
            <div class="border-t border-slate-700/50 pt-2 mt-2">
-                {% for d in c.drafts[:5] %}
+                {% for d in c.drafts[:3] %}
                <div class="text-xs truncate mb-0.5" title="{{ d.name }}: {{ d.title }}">
                    <a href="/drafts/{{ d.name }}" class="text-blue-400/70 hover:text-blue-300 transition" onclick="event.stopPropagation()">{{ d.title }}</a>
                </div>
                {% endfor %}
-                {% if c.draft_count > 5 %}
+                {% if c.draft_count > 3 %}
-                <div class="text-xs text-slate-600 mt-1">+{{ c.draft_count - 5 }} more drafts</div>
+                <div class="text-xs text-slate-600 mt-1">+{{ c.draft_count - 3 }} more drafts</div>
                {% endif %}
            </div>
            {% endif %}
            <!-- Expanded detail (hidden by default) -->
            <div class="cluster-detail hidden mt-3 border-t border-slate-700/50 pt-3" id="authorCluster-{{ c.id }}">
                <!-- All members with org -->
                <h4 class="text-xs font-semibold text-slate-300 mb-2 uppercase tracking-wide">All {{ c.size }} Authors</h4>
                <div class="max-h-48 overflow-y-auto mb-3 space-y-1">
                    {% for member in c.members %}
                    <div class="text-xs flex items-center justify-between gap-2">
                        <a href="/drafts?q={{ member | urlencode }}" class="text-slate-300 hover:text-blue-400 transition truncate" onclick="event.stopPropagation()">{{ member }}</a>
                        {% set member_org = c.member_orgs[member] if c.member_orgs is defined and member in c.member_orgs else '' %}
                        {% if member_org %}
                        <span class="text-slate-600 text-[10px] truncate max-w-[120px] flex-shrink-0">{{ member_org }}</span>
                        {% endif %}
                    </div>
                    {% endfor %}
                </div>
                <!-- All drafts -->
                {% if c.drafts %}
                <h4 class="text-xs font-semibold text-slate-300 mb-2 uppercase tracking-wide">All {{ c.draft_count }} Drafts</h4>
                <div class="max-h-48 overflow-y-auto space-y-1.5">
                    {% for d in c.drafts %}
                    <div class="text-xs" title="{{ d.name }}">
                        <a href="/drafts/{{ d.name }}" class="text-blue-400/70 hover:text-blue-300 transition" onclick="event.stopPropagation()">{{ d.title }}</a>
                        <span class="text-slate-600 font-mono text-[10px] ml-1">{{ d.name | replace('draft-', '') | truncate(25) }}</span>
                    </div>
                    {% endfor %}
                </div>
                {% endif %}
            </div>
        </div>
        {% endfor %}
    </div>
@@ -478,6 +516,20 @@ const network = {{ network | tojson }};
        });
    });
    // Toggle expand/collapse on cluster card chevron click
    document.querySelectorAll('.cluster-card').forEach(card => {
        card.addEventListener('click', function(e) {
            // Don't toggle if clicking a link or the highlight header
            if (e.target.closest('a') || e.target.closest('[onclick*="highlightCluster"]')) return;
            const detail = card.querySelector('.cluster-detail');
            const chevron = card.querySelector('.cluster-chevron');
            if (detail) {
                detail.classList.toggle('hidden');
                chevron.style.transform = detail.classList.contains('hidden') ? '' : 'rotate(180deg)';
            }
        });
    });
    // Expose cluster highlighting globally
    window.highlightCluster = function(clusterId) {
        const cluster = (network.clusters || []).find(c => c.id === clusterId);
--- a/src/webui/templates/comparison.html
+++ b/src/webui/templates/comparison.html
@@ -71,7 +71,7 @@
            {{ draft.title }}
        </a>
        <div class="text-xs text-slate-600 font-mono mt-1">{{ draft.name }}</div>
-        <div class="text-xs text-slate-500 mt-2 line-clamp-3">{{ draft.abstract[:200] }}</div>
+        <div class="text-xs text-slate-500 mt-2 line-clamp-3">{{ (draft.abstract | striptags)[:200] }}</div>
        {% if draft.rating %}
        <!-- Rating radar -->
@@ -79,9 +79,15 @@
            {% for dim, label in [('novelty', 'Nov'), ('maturity', 'Mat'), ('relevance', 'Rel'), ('momentum', 'Mom'), ('overlap', 'Ovl')] %}
            <div>
                <div class="text-xs text-slate-500">{{ label }}</div>
                {% if dim == 'overlap' %}
                <div class="text-sm font-semibold {% if draft.rating[dim] <= 2 %}text-green-400{% elif draft.rating[dim] <= 3 %}text-yellow-400{% else %}text-red-400{% endif %}">
                    {{ draft.rating[dim] }}
                </div>
                {% else %}
                <div class="text-sm font-semibold {% if draft.rating[dim] >= 4 %}text-green-400{% elif draft.rating[dim] >= 3 %}text-yellow-400{% else %}text-red-400{% endif %}">
                    {{ draft.rating[dim] }}
                </div>
                {% endif %}
            </div>
            {% endfor %}
        </div>
--- a/src/webui/templates/draft_detail.html
+++ b/src/webui/templates/draft_detail.html
@@ -91,7 +91,7 @@
                <svg class="w-4 h-4 text-slate-500" fill="none" stroke="currentColor" viewBox="0 0 24 24"><path stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M4 6h16M4 12h16M4 18h7"/></svg>
                Abstract
            </h2>
-            <p class="text-sm text-slate-400 leading-relaxed">{{ draft.abstract or "No abstract available." }}</p>
+            <p class="text-sm text-slate-400 leading-relaxed">{{ (draft.abstract | striptags) or "No abstract available." }}</p>
        </div>
        <!-- Rating Analysis -->
@@ -120,10 +120,18 @@
                            <svg class="w-3.5 h-3.5 text-slate-500" fill="none" stroke="currentColor" viewBox="0 0 24 24"><path stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="{{ icon }}"/></svg>
                            <span class="text-xs font-semibold text-slate-300 uppercase tracking-wide">{{ label }}</span>
                        </div>
                        {% if dim == "overlap" %}
                        <span class="text-lg font-bold {% if val <= 2 %}text-green-400{% elif val <= 3 %}text-amber-400{% else %}text-red-400{% endif %}">{{ val }}<span class="text-xs text-slate-600 font-normal">/5</span></span>
                        {% else %}
                        <span class="text-lg font-bold {% if val >= 4 %}text-green-400{% elif val >= 3 %}text-amber-400{% else %}text-red-400{% endif %}">{{ val }}<span class="text-xs text-slate-600 font-normal">/5</span></span>
                        {% endif %}
                    </div>
                    <div class="dim-progress mb-2">
                        {% if dim == "overlap" %}
                        <div class="dim-progress-fill {% if val <= 2 %}dim-high{% elif val <= 3 %}dim-mid{% else %}dim-low{% endif %}" style="width: {{ val * 20 }}%"></div>
                        {% else %}
                        <div class="dim-progress-fill {% if val >= 4 %}dim-high{% elif val >= 3 %}dim-mid{% else %}dim-low{% endif %}" style="width: {{ val * 20 }}%"></div>
                        {% endif %}
                    </div>
                    {% if draft.rating[dim + '_note'] %}
                    <p class="text-xs text-slate-500 leading-relaxed">{{ draft.rating[dim + '_note'] }}</p>
@@ -231,7 +239,11 @@
                {% for dim, abbr in [("novelty","N"), ("maturity","M"), ("overlap","O"), ("momentum","Mo"), ("relevance","R")] %}
                {% set v = draft.rating[dim] %}
                <div>
                    {% if dim == "overlap" %}
                    <div class="text-xs font-bold {% if v <= 2 %}text-green-400{% elif v <= 3 %}text-amber-400{% else %}text-red-400{% endif %}">{{ v }}</div>
                    {% else %}
                    <div class="text-xs font-bold {% if v >= 4 %}text-green-400{% elif v >= 3 %}text-amber-400{% else %}text-red-400{% endif %}">{{ v }}</div>
                    {% endif %}
                    <div class="text-[9px] text-slate-600 uppercase">{{ abbr }}</div>
                </div>
                {% endfor %}
--- a/src/webui/templates/drafts.html
+++ b/src/webui/templates/drafts.html
@@ -68,6 +68,11 @@
        color: #c084fc;
        border: 1px solid rgba(168, 85, 247, 0.3);
    }
    .source-nist {
        background: rgba(6, 182, 212, 0.15);
        color: #22d3ee;
        border: 1px solid rgba(6, 182, 212, 0.3);
    }
    .source-generated {
        background: rgba(148, 163, 184, 0.15);
        color: #94a3b8;
@@ -180,6 +185,7 @@
                    <option value="etsi" {% if current_source == 'etsi' %}selected{% endif %}>ETSI</option>
                    <option value="itu" {% if current_source == 'itu' %}selected{% endif %}>ITU-T</option>
                    <option value="iso" {% if current_source == 'iso' %}selected{% endif %}>ISO/IEC</option>
                    <option value="nist" {% if current_source == 'nist' %}selected{% endif %}>NIST</option>
                </select>
            </div>
            <!-- Sort -->
@@ -426,7 +432,7 @@
                    <td class="px-4 py-3 hidden xl:table-cell">
                        <div class="flex items-center gap-1.5">
                            <span class="dim-bar-bg">
-                                <span class="dim-bar-fill {% if d.overlap >= 4 %}dim-fill-high{% elif d.overlap >= 3 %}dim-fill-mid{% else %}dim-fill-low{% endif %}"
+                                <span class="dim-bar-fill {% if d.overlap <= 2 %}dim-fill-high{% elif d.overlap <= 3 %}dim-fill-mid{% else %}dim-fill-low{% endif %}"
                                      style="width: {{ (d.overlap / 5 * 100)|int }}%"></span>
                            </span>
                            <span class="text-xs text-slate-500 font-mono w-4 text-right">{{ d.overlap }}</span>
--- a/src/webui/templates/idea_clusters.html
+++ b/src/webui/templates/idea_clusters.html
@@ -63,6 +63,20 @@
        <div id="treemapPlot" style="height: 450px;"></div>
    </div>
    <!-- Cluster relationship network -->
    <div id="networkSection" class="bg-slate-900 rounded-xl border border-slate-800 p-5 mb-6 hidden">
        <h2 class="text-sm font-semibold text-slate-300 mb-1">Cluster Relationships</h2>
        <p class="text-xs text-slate-500 mb-3">Network showing how idea clusters relate to each other. Thicker lines = stronger semantic similarity. Click a link to see the connecting ideas.</p>
        <div id="networkPlot" style="height: 560px;"></div>
        <div id="linkDetail" class="hidden mt-4 bg-slate-800/50 rounded-lg p-4 border border-slate-700/50">
            <div class="flex items-center justify-between mb-2">
                <h3 class="text-sm font-semibold text-white" id="linkTitle"></h3>
                <button onclick="document.getElementById('linkDetail').classList.add('hidden')" class="text-slate-500 hover:text-white text-xs">✕</button>
            </div>
            <div id="linkContent" class="text-xs text-slate-400 space-y-2"></div>
        </div>
    </div>
    <!-- Cluster cards grid -->
    <h2 class="text-lg font-semibold text-white mb-4">Cluster Details</h2>
    <div id="clusterGrid" class="grid grid-cols-1 md:grid-cols-2 xl:grid-cols-3 gap-4 mb-6">
@@ -180,6 +194,135 @@ if (data.empty) {
        }, CFG);
    }
    // --- Cluster Relationship Network ---
    const links = data.links || [];
    if (links.length > 0) {
        document.getElementById('networkSection').classList.remove('hidden');
        // Build node set from clusters that have links
        const linkedIds = new Set();
        links.forEach(l => { linkedIds.add(l.source); linkedIds.add(l.target); });
        const nodes = data.clusters.filter(c => linkedIds.has(c.id));
        const nodeMap = {};
        nodes.forEach((n, i) => { nodeMap[n.id] = i; });
        // Force-directed layout using Plotly scatter + annotations for edges
        // Position nodes in a circle, then use link structure
        const n = nodes.length;
        const nodeX = nodes.map((_, i) => Math.cos(2 * Math.PI * i / n) * 4);
        const nodeY = nodes.map((_, i) => Math.sin(2 * Math.PI * i / n) * 4);
        // Simple force-directed: pull linked nodes closer
        for (let iter = 0; iter < 80; iter++) {
            for (const link of links) {
                const si = nodeMap[link.source];
                const ti = nodeMap[link.target];
                if (si === undefined || ti === undefined) continue;
                const dx = nodeX[ti] - nodeX[si];
                const dy = nodeY[ti] - nodeY[si];
                const dist = Math.sqrt(dx*dx + dy*dy) || 1;
                const force = (link.best_pair_sim - 0.5) * 0.15;
                nodeX[si] += dx/dist * force;
                nodeY[si] += dy/dist * force;
                nodeX[ti] -= dx/dist * force;
                nodeY[ti] -= dy/dist * force;
            }
            // Repulsion between all nodes
            for (let i = 0; i < n; i++) {
                for (let j = i+1; j < n; j++) {
                    const dx = nodeX[j] - nodeX[i];
                    const dy = nodeY[j] - nodeY[i];
                    const dist = Math.sqrt(dx*dx + dy*dy) || 0.1;
                    if (dist < 1.5) {
                        const repel = 0.3 / (dist * dist);
                        nodeX[i] -= dx/dist * repel;
                        nodeY[i] -= dy/dist * repel;
                        nodeX[j] += dx/dist * repel;
                        nodeY[j] += dy/dist * repel;
                    }
                }
            }
        }
        // Edge traces (one per link for click handling)
        const edgeTraces = links.map((link, li) => {
            const si = nodeMap[link.source];
            const ti = nodeMap[link.target];
            if (si === undefined || ti === undefined) return null;
            const width = 1 + (link.best_pair_sim - 0.5) * 8;
            const opacity = 0.3 + (link.best_pair_sim - 0.5) * 1.2;
            return {
                x: [nodeX[si], nodeX[ti], null],
                y: [nodeY[si], nodeY[ti], null],
                mode: 'lines',
                line: { width: width, color: `rgba(100,116,139,${opacity})` },
                hoverinfo: 'text',
                text: `${link.source_theme} ↔ ${link.target_theme}<br>Similarity: ${(link.best_pair_sim * 100).toFixed(0)}%`,
                customdata: [li, li, null],
                showlegend: false,
            };
        }).filter(Boolean);
        // Node trace
        const nodeTrace = {
            x: nodeX, y: nodeY,
            mode: 'markers+text',
            type: 'scatter',
            marker: {
                size: nodes.map(n => 12 + Math.sqrt(n.size) * 3),
                color: nodes.map((_, i) => PALETTE[nodes[i].id % PALETTE.length]),
                line: { width: 2, color: 'rgba(15,23,42,0.8)' },
            },
            text: nodes.map(n => n.theme.length > 25 ? n.theme.substring(0, 22) + '...' : n.theme),
            textposition: 'top center',
            textfont: { size: 10, color: '#cbd5e1' },
            hovertext: nodes.map(n =>
                `<b>${n.theme}</b><br>${n.size} ideas, ${n.drafts.length} drafts`
            ),
            hoverinfo: 'text',
            showlegend: false,
        };
        Plotly.newPlot('networkPlot', [...edgeTraces, nodeTrace], {
            ...PLOTLY_LAYOUT,
            xaxis: { visible: false, showgrid: false, zeroline: false },
            yaxis: { visible: false, showgrid: false, zeroline: false },
            hovermode: 'closest',
            margin: { t: 10, r: 20, b: 10, l: 20 },
        }, CFG);
        // Click handler for edges — show link detail
        document.getElementById('networkPlot').on('plotly_click', function(ev) {
            const pt = ev.points[0];
            if (pt.data.customdata && pt.data.customdata[pt.pointNumber] !== null) {
                const link = links[pt.data.customdata[pt.pointNumber]];
                if (!link) return;
                const detail = document.getElementById('linkDetail');
                const simPct = (link.best_pair_sim * 100).toFixed(0);
                document.getElementById('linkTitle').innerHTML =
                    `<span style="color:${PALETTE[link.source % PALETTE.length]}">${link.source_theme}</span>` +
                    ` <span class="text-slate-500">↔</span> ` +
                    `<span style="color:${PALETTE[link.target % PALETTE.length]}">${link.target_theme}</span>` +
                    ` <span class="text-slate-500 text-xs font-normal ml-2">${simPct}% similar</span>`;
                document.getElementById('linkContent').innerHTML = `
                    <div class="grid grid-cols-2 gap-4">
                        <div class="bg-slate-900/50 rounded p-3 border border-slate-700/30">
                            <div class="text-slate-300 font-medium mb-1">${link.idea_a}</div>
                            <a href="/drafts/${link.idea_a_draft}" class="text-blue-400/70 hover:text-blue-300 text-[10px] font-mono">${link.idea_a_draft}</a>
                        </div>
                        <div class="bg-slate-900/50 rounded p-3 border border-slate-700/30">
                            <div class="text-slate-300 font-medium mb-1">${link.idea_b}</div>
                            <a href="/drafts/${link.idea_b_draft}" class="text-blue-400/70 hover:text-blue-300 text-[10px] font-mono">${link.idea_b_draft}</a>
                        </div>
                    </div>
                    <p class="text-slate-500 text-[10px] mt-1">These two ideas from different clusters have the strongest cross-cluster similarity.</p>
                `;
                detail.classList.remove('hidden');
            }
        });
    }
    // --- Cluster Cards ---
    const grid = document.getElementById('clusterGrid');
@@ -190,15 +333,42 @@ if (data.empty) {
            if (filter === 'large' && cluster.size < 10) return;
            const color = PALETTE[i % PALETTE.length];
-            const topIdeas = cluster.ideas.slice(0, 5);
+            const cardId = `cluster-${i}`;
-            const ideaListHtml = topIdeas.map(idea =>
+            const topIdeas = cluster.ideas.slice(0, 3);
-                `<li class="text-xs text-slate-400 truncate" title="${idea.description || idea.title}">
+
-                    <span class="text-slate-300">${idea.title}</span>
+            // Deduplicate ideas by title, track which drafts have each
-                </li>`
+            const ideaByTitle = {};
-            ).join('');
+            cluster.ideas.forEach(idea => {
-            const extraCount = cluster.size - topIdeas.length;
+                if (!ideaByTitle[idea.title]) {
-            const extraHtml = extraCount > 0
+                    ideaByTitle[idea.title] = { ...idea, drafts: [] };
-                ? `<li class="text-xs text-slate-600">+${extraCount} more</li>` : '';
+                }
                ideaByTitle[idea.title].drafts.push(idea.draft_name);
            });
            const uniqueIdeas = Object.values(ideaByTitle);
            // Preview: first 3 unique ideas
            const previewHtml = uniqueIdeas.slice(0, 3).map(idea => {
                const draftTag = idea.drafts.length > 1
                    ? `<span class="text-slate-600">(${idea.drafts.length} drafts)</span>`
                    : `<span class="text-slate-600">${idea.drafts[0].replace('draft-', '').substring(0, 20)}</span>`;
                return `<li class="text-xs text-slate-400 truncate" title="${idea.description || idea.title}">
                    <span class="text-slate-300">${idea.title}</span> ${draftTag}
                </li>`;
            }).join('');
            const previewExtra = uniqueIdeas.length > 3
                ? `<li class="text-xs text-slate-600">+${uniqueIdeas.length - 3} more unique ideas</li>` : '';
            // Full idea list (shown on expand)
            const fullIdeasHtml = uniqueIdeas.map(idea => {
                const draftLinks = idea.drafts.map(d =>
                    `<a href="/drafts/${d}" class="text-blue-400/70 hover:text-blue-300 transition">${d.replace('draft-', '').substring(0, 28)}</a>`
                ).join(', ');
                return `<div class="py-2 border-b border-slate-800/50 last:border-0">
                    <div class="text-xs text-slate-200 font-medium">${idea.title}</div>
                    ${idea.description ? `<div class="text-xs text-slate-500 mt-0.5 leading-relaxed">${idea.description.substring(0, 200)}</div>` : ''}
                    <div class="text-[10px] text-slate-600 mt-1 font-mono">${draftLinks}</div>
                </div>`;
            }).join('');
            // WG badges
            const wgBadges = (cluster.wgs || []).filter(w => w.wg !== 'none').map(w =>
@@ -224,22 +394,39 @@ if (data.empty) {
                ? `<span class="text-xs bg-amber-900/30 text-amber-400 px-1.5 py-0.5 rounded">cross-WG</span>` : '';
            const card = document.createElement('div');
-            card.className = 'bg-slate-900 rounded-xl border p-5 ' +
+            card.className = 'bg-slate-900 rounded-xl border p-5 cursor-pointer hover:border-slate-600 transition ' +
                (cluster.cross_wg ? 'border-amber-800/40' : 'border-slate-800');
            card.onclick = () => {
                const detail = document.getElementById(cardId);
                const chevron = document.getElementById(`chevron-${i}`);
                if (detail.classList.contains('hidden')) {
                    detail.classList.remove('hidden');
                    chevron.style.transform = 'rotate(180deg)';
                } else {
                    detail.classList.add('hidden');
                    chevron.style.transform = '';
                }
            };
            card.innerHTML = `
                <div class="flex items-center gap-2 mb-3">
                    <div class="w-3 h-3 rounded-full flex-shrink-0" style="background: ${color}"></div>
                    <h3 class="text-sm font-semibold text-white truncate">${cluster.theme}</h3>
                    ${crossBadge}
                    <span class="ml-auto text-xs text-slate-500 flex-shrink-0">${cluster.size} ideas</span>
                    <svg id="chevron-${i}" class="w-4 h-4 text-slate-500 flex-shrink-0 transition-transform" fill="none" stroke="currentColor" viewBox="0 0 24 24"><path stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M19 9l-7 7-7-7"/></svg>
                </div>
-                <ul class="space-y-1 mb-3">${ideaListHtml}${extraHtml}</ul>
+                <ul class="space-y-1 mb-3">${previewHtml}${previewExtra}</ul>
                ${(wgBadges || noneHtml) ? `<div class="mb-2"><p class="text-xs text-slate-500 mb-1">Working Groups</p><div class="flex flex-wrap gap-1">${wgBadges} ${noneHtml}</div></div>` : ''}
                ${catBadges ? `<div class="mb-2"><p class="text-xs text-slate-500 mb-1">Categories</p><div class="flex flex-wrap gap-1">${catBadges}</div></div>` : ''}
                <div class="border-t border-slate-800 pt-3">
                    <p class="text-xs text-slate-500 mb-1">${cluster.drafts.length} source draft${cluster.drafts.length !== 1 ? 's' : ''}</p>
                    <div class="flex flex-wrap gap-1">${draftBadges}${extraDrafts}</div>
                </div>
                <!-- Expanded detail (hidden by default) -->
                <div id="${cardId}" class="hidden mt-4 border-t border-slate-700 pt-4">
                    <h4 class="text-xs font-semibold text-slate-300 mb-2 uppercase tracking-wide">All ${uniqueIdeas.length} unique ideas</h4>
                    <div class="max-h-80 overflow-y-auto pr-1">${fullIdeasHtml}</div>
                </div>
            `;
            grid.appendChild(card);
        });
--- a/src/webui/templates/ratings.html
+++ b/src/webui/templates/ratings.html
@@ -185,8 +185,13 @@ document.getElementById('scatter').on('plotly_click', function(data) {
        return 'score-low';
    }
-    function dimBadge(val) {
+    function dimBadge(val, inverted = false) {
-        const cls = val >= 4 ? 'text-green-400' : val >= 3 ? 'text-yellow-400' : 'text-slate-500';
+        let cls;
        if (inverted) {
            cls = val <= 2 ? 'text-green-400' : val <= 3 ? 'text-yellow-400' : 'text-red-400';
        } else {
            cls = val >= 4 ? 'text-green-400' : val >= 3 ? 'text-yellow-400' : 'text-slate-500';
        }
        return `<span class="${cls}">${val}</span>`;
    }
@@ -207,7 +212,7 @@ document.getElementById('scatter').on('plotly_click', function(data) {
            <td class="px-4 py-3 text-center">${dimBadge(d.maturity)}</td>
            <td class="px-4 py-3 text-center">${dimBadge(d.relevance)}</td>
            <td class="px-4 py-3 text-center">${dimBadge(d.momentum)}</td>
-            <td class="px-4 py-3 text-center">${dimBadge(d.overlap)}</td>
+            <td class="px-4 py-3 text-center">${dimBadge(d.overlap, true)}</td>
            <td class="px-4 py-3">
                <span class="px-2 py-0.5 rounded text-[10px] bg-slate-800 text-slate-400">${d.category}</span>
            </td>