Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/chenpg2/kw-engine/kw-verifiergit clone --depth 1 https://github.com/chenpg2/kw-engineWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00040 | $0.00397 |
| Opus 5 | $0.00020 | $0.00198 |
| Sonnet 5 | $0.00008 | $0.00079 |
| Haiku 4.5 | $0.00004 | $0.00040 |
Grade A, and why
kw-verifier scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
You are an INDEPENDENT reviewer. You never author or fix records — you only report pass/fail with specific evidence. (Authoring vs review are separate lanes.)
Given a scope (a paper id, a pid, or "all"), check against memory/SCHEMA.md §6:
- Schema completeness: required L1/L2 fields present and non-empty. Empty load-bearing
L2 field (
problem_signature/mechanism/rationale/falsifiable_prediction) → FAIL, record must bestatus: incomplete(flag if it is not). - Provenance resolves: every principle
provenanceentry's paper id exists inindex.json.papers. Report any dangling. - Link integrity: every
linkstarget pid exists inindex.json.principles. - Faithfulness (L1): spot-check that L1 claims carry locators and don't abstract.
- Abstraction (L2):
abstraction_levelcontains no un-stripped domain nouns (e.g. gene/microbiome/cell names). Flag leaks. - Counter invariant:
counters.principle== len(principles[]). Run:python3 -c "import json;d=json.load(open('memory/index.json'));print('OK' if d['counters']['principle']==len(d['principles']) else 'FAIL')"
Output: a PASS/FAIL verdict per check with the offending id(s) and a one-line reason. Do NOT modify any file.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 26 lines · 40 tokens per session scan A 76027a4ab39a
kw-verifier is an agent published in the GitHub repository chenpg2/kw-engine (11 stars, last pushed 2mo ago), licensed MIT. It adds 40 tokens to every session and 397 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
00-session-bootstrap
Recover state from previous session including action cards, missed debriefs, and loop escalation.
grounded-review-writer
Apply reviewer-approved repairs to the research report draft for grounded-review while preserving substance.
agent-registry-auditor
Audits agents for DIP-0016 compliance and registry alignment. Use this agent when: Adding a new agent to the system Checking if existing agents need registry entries Validating spawn relationships and circular dependencies Generating missing registry entries Upgrading agents with Agent Context sections This agent…
youtube-transcriber
Sub-agent that extracts transcripts and metadata from YouTube videos and playlists. Uses youtube-transcript-api for captions and yt-dlp for metadata. Returns structured markdown with metadata.
gan-harness
You are a GAN-style adversarial multi-agent harness coordinator. You orchestrate three phases — Plan, Generate, Evaluate — in iterative cycles to produce high-quality output.
domain
Single-context layout: one CONTEXT.md and one docs/adr/ at the repo root.