Fix remaining critical, high, and medium issues from 4-perspective review
Critical fixes:
- Fix rating clamp range 1-10 → 1-5 (actual scale)
- Add `ietf ideas convergence` command (SequenceMatcher at 0.75 threshold)
- Fix "628 cross-org ideas" → 130 (verified from current DB) across 8 files
Security fixes:
- Sanitize FTS5 query input (strip special chars + boolean operators)
- Add rate limiting (10 req/min/IP) on Claude-calling endpoints
- Change <path:name> → <string:name> on draft routes
Codebase fixes:
- Add Database context manager (__enter__/__exit__)
- Wire false_positive filtering into queries (exclude by default in web UI)
- Fix Post 3 arithmetic ("~300" → "~409" distinct proposals)
Content & licensing:
- Add MIT LICENSE file
- Add IPR/FRAND notes (BCP 79, RFC 8179) to Posts 03 and 07
- Qualify "4:1 safety ratio" with monthly variation in 6 remaining files
- Add "Data as of March 2026" freeze-date headers to all 10 blog posts
- Hedge causal language in Post 04
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
@@ -51,6 +51,72 @@
|
||||
</div>
|
||||
</div>
|
||||
|
||||
<div class="bg-slate-900 rounded-xl border border-slate-800 p-6 mb-6">
|
||||
<h2 class="text-lg font-semibold text-white mb-3">Data Collection Methodology</h2>
|
||||
<p class="text-sm text-slate-400 leading-relaxed mb-4">
|
||||
Drafts are discovered by searching the
|
||||
<a href="https://datatracker.ietf.org" class="text-blue-400 hover:text-blue-300 transition">IETF Datatracker API</a>
|
||||
for documents whose abstract contains any of the following keywords.
|
||||
Only drafts submitted since <span class="text-slate-200 font-medium">{{ fetch_since }}</span> are included.
|
||||
</p>
|
||||
|
||||
<h3 class="text-sm font-semibold text-slate-300 mb-2">Search Keywords</h3>
|
||||
<div class="flex flex-wrap gap-2 mb-4">
|
||||
{% for kw in search_keywords %}
|
||||
<span class="px-2.5 py-1 bg-blue-500/10 text-blue-400 border border-blue-500/20 rounded-md text-xs font-mono">{{ kw }}</span>
|
||||
{% endfor %}
|
||||
</div>
|
||||
|
||||
<h3 class="text-sm font-semibold text-slate-300 mb-2">Analysis Pipeline</h3>
|
||||
<div class="text-sm text-slate-400 leading-relaxed space-y-2">
|
||||
<p><span class="text-slate-200 font-medium">1. Fetch</span> — Query Datatracker API for each keyword, deduplicate by draft name, download full text.</p>
|
||||
<p><span class="text-slate-200 font-medium">2. Rate</span> — Claude rates each draft on 5 dimensions (novelty, maturity, overlap, momentum, relevance) from 1–5, with per-dimension explanations.</p>
|
||||
<p><span class="text-slate-200 font-medium">3. Categorize</span> — Claude assigns one or more topic categories (e.g., "A2A protocols", "Agent identity/auth").</p>
|
||||
<p><span class="text-slate-200 font-medium">4. Extract Ideas</span> — Claude extracts distinct technical ideas from each draft, with novelty scores.</p>
|
||||
<p><span class="text-slate-200 font-medium">5. Embed</span> — Ollama generates vector embeddings for similarity analysis and clustering.</p>
|
||||
<p><span class="text-slate-200 font-medium">6. Author Network</span> — Author and affiliation data fetched from Datatracker to build collaboration graphs.</p>
|
||||
<p><span class="text-slate-200 font-medium">7. Gap Analysis</span> — Claude identifies areas where no existing draft adequately addresses a need.</p>
|
||||
</div>
|
||||
|
||||
<div class="mt-4 p-3 bg-slate-800/50 rounded-lg border border-slate-700/50">
|
||||
<p class="text-xs text-slate-500">
|
||||
<span class="text-amber-400/70 font-semibold">Note on keyword selection:</span>
|
||||
Keywords determine which drafts are included. Broad terms like "agent" and "autonomous" cast a wide net
|
||||
(catching some tangentially related drafts), while specific terms like "ai-agent" and "agentic" target
|
||||
the core AI agent space. The false-positive flag in ratings helps filter out irrelevant matches.
|
||||
Suggestions for additional keywords are welcome.
|
||||
</p>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
<div class="bg-slate-900 rounded-xl border border-slate-800 p-6 mb-6">
|
||||
<h2 class="text-lg font-semibold text-white mb-3">Scoring Methodology</h2>
|
||||
<div class="text-sm text-slate-400 leading-relaxed space-y-3">
|
||||
<p>Each draft is rated by Claude AI on five dimensions, scored from 1 (lowest) to 5 (highest):</p>
|
||||
<div class="overflow-x-auto">
|
||||
<table class="w-full text-sm">
|
||||
<thead>
|
||||
<tr class="border-b border-slate-700 text-left text-xs text-slate-500">
|
||||
<th class="py-2 pr-4 font-medium">Dimension</th>
|
||||
<th class="py-2 font-medium">What it measures</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody class="divide-y divide-slate-800/50">
|
||||
<tr><td class="py-2 pr-4 text-slate-300 font-medium">Novelty</td><td class="py-2">Originality of contribution. Does it introduce genuinely new ideas?</td></tr>
|
||||
<tr><td class="py-2 pr-4 text-slate-300 font-medium">Maturity</td><td class="py-2">Completeness of the specification. Ready for implementation?</td></tr>
|
||||
<tr><td class="py-2 pr-4 text-slate-300 font-medium">Overlap</td><td class="py-2">Duplication with other drafts. High = redundant. <em>Inverted in composite score.</em></td></tr>
|
||||
<tr><td class="py-2 pr-4 text-slate-300 font-medium">Momentum</td><td class="py-2">Activity level. Revisions, WG adoption, multi-org authorship.</td></tr>
|
||||
<tr><td class="py-2 pr-4 text-slate-300 font-medium">Relevance</td><td class="py-2">How directly related to AI agent infrastructure.</td></tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</div>
|
||||
<p class="mt-2">
|
||||
<span class="text-slate-200 font-medium">Composite score</span> = (novelty + maturity + (5 - overlap) + momentum + relevance) / 5.
|
||||
Overlap is inverted so lower overlap contributes positively.
|
||||
</p>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
<div class="bg-slate-900 rounded-xl border border-slate-800 p-6">
|
||||
<h2 class="text-lg font-semibold text-white mb-3">Tech Stack</h2>
|
||||
<ul class="text-sm text-slate-400 space-y-2">
|
||||
|
||||
Reference in New Issue
Block a user