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 skills/tpierrain/kenjaku/mutation-testingnpx skills add tpierrain/kenjaku --skill mutation-testinggit clone --depth 1 https://github.com/tpierrain/kenjakuWhat 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 | $0.00119 | $0.02669 |
| Opus 5 | $0.00060 | $0.01334 |
| Sonnet 5 | $0.00024 | $0.00534 |
| Haiku 4.5 | $0.00012 | $0.00267 |
Grade A, and why
mutation-testing 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 3d ago.
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.
How it starts
The opening of the file, as written. The whole thing — 166 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Mutation testing, on this repo
The mutation score is the judge of the tests, not a ceremony (test-first-discipline). This skill
is the judgement half of that: when a pass is due, how to read what it gives back, and what to
write down. The mechanics are the other half and they are a script —
maintainers/mutation/mutate-one.mjs — because every mechanical trap
below has already cost a real run at least once, and a written reflex was measured as insufficient
(CONVENTIONS §5quinquies).
Not here: how to write a killing assertion. That is
test-first-discipline(a matcher on everythrows, assert the whole object, triangulate bounds, collections ≥2 unsorted, a double's answer as a fingerprint). This skill starts once the run has produced a number.Nor the portable version of the traps below. Since 2026-08-20,
test-first-discipline§"A mutation run LIES to you" carries the five ways any run hands back a number that measures nothing, and the survivor triage that precedes writing a test — tool- and language-agnostic. This file owns the local answer: which command, which worktree, which package. Read the doctrine there, run it here.
1. When a pass is due — and on WHAT SURFACE
🔑 The scope of a run is the scope of the change. A new file is measured whole; an existing one is measured by its changed lines (
"scripts/file.mjs:147-160"— the runner passes the range straight through to Stryker). Whole-file re-runs of large existing files are where this practice dies of slowness: measured on the S6 block, one cost 7 minutes to return a survivor list byte-identical to the previous run, and a second was killed by the 10-minute cap. The same real defect, re-measured on its hunk, died in 44 seconds (17 mutants instead of 396). The reasoning and the numbers live intest-first-discipline§ When to run it, and on what; the commands live in CONVENTIONS §5quinquies.The bonus is in §3: on a hunk-scoped run every survivor is yours, so there is no score drift to attribute and no survivor list to diff against last time.
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.
- 3d ago First seen · 166 lines · 119 tokens per session scan A 8d4b76943413
mutation-testing is a skill published in the GitHub repository tpierrain/kenjaku (80 stars, last pushed 9d ago), licensed Apache-2.0. It adds 119 tokens to every session and 2,669 once invoked, about $0.0006 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 skills, from other repositories
lint
Run the LLM Wiki health check: orphans, broken links, contradictions, stale pages, index sync, MOC coverage, frontmatter fields, Exploration Gate, v5 verification fields, Bias Check, cross-vault links, and Core Context freshness; also responds to /lint and 위키 점검.
refresh-context
Re-snapshot Core Context.md from the 9 mothership system files and key essays when the mothership changes or the snapshot is older than 30 days; also responds to /refresh-context and 코어 컨텍스트 갱신.
status
Show the LLM Wiki state at a glance: counts, recent activity, index drift, inbox backlog, collectionPurpose/mainVault/explored/verificationStatus coverage, and Core Context snapshot age; also responds to /status and 상태 점검.
capture-tabs
Capture Chrome tab groups, browser research bundles, or ChatGPT/Gemini/Grok/Claude/Perplexity sessions into LLM Wiki Inbox markdown; also responds to 탭 캡처, 리서치 캡처, Deep Research 캡처, and /capture-tabs.
ingest
Ingest a URL, file, raw text, or Inbox item into the LLM Wiki with the purpose gate, mothership link search, raw-source preservation, Wiki compilation, index/log update, and review; also responds to /ingest and 인제스트.
reindex
Manually re-run qmd indexing and embeddings for wiki, rawsources, and queries after Obsidian edits, bulk moves, or embedding-model changes; also responds to /reindex and 재색인.