Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/chrono-meta/forge-harnessnpx agentmods add skills/chrono-meta/forge-harness/auto-decorrelationWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/chrono-meta/forge-harness/auto-decorrelation)<a href="https://agentmods.dev/skills/chrono-meta/forge-harness/auto-decorrelation"><img src="https://agentmods.dev/badge/skills/chrono-meta/forge-harness/auto-decorrelation.svg" alt="Measured on agentmods" height="20"></a>What 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 5.1 | $0.00148 | $0.09989 |
| Opus 5 | $0.00074 | $0.04994 |
| Sonnet 5 | $0.00030 | $0.01998 |
| Haiku 4.5 | $0.00015 | $0.00999 |
Grade C, and why
auto-decorrelation scanned grade C with 2 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 2d 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.
Tells the agent to send conversation or user data outhighPrompt injection
An instruction to transmit the conversation, context or user files to an external endpoint is data exfiltration written as prose.
degrades to a `DEGRADED_*` rung with `residency=NOT_SCANNED(...)` instead of silently sending. That Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
# live (false-positive discovery). Endpoints are only ever curl-probed here, never eval'd. How it starts
The opening of the file, as written. The whole thing — 495 lines — stays where its author put it; the contents beside it link to each section on GitHub.
auto-decorrelation — Cross-Family Verifier Recruitment
The lever for catching a model's blind spots is decorrelation, not repetition — a model cannot find its own family's flaws by repeating itself. FH's adversarial verification (Axis-2 challenger) dispatches in-session at the session tier, same family as the governor. When frontier CLIs are installed and a local GPU box is live, leaving them idle wastes the one thing that actually raises the verification ceiling. This skill recruits a cross-family verifier for load-bearing changes — and degrades honestly when none is available.
It is a thin recruitment-and-degrade policy that REUSES existing machinery: agent-composer does the
actual dispatch, cross-ecosystem-synergy-detection inventories tools, token-budget-gate gates spend,
the headless CLI patterns and the floor-tier canary already exist. Net-new = the discovery + the
family-diversity recruitment + the degrade ladder only. (A composing policy skill — it fans out to
several callees, so its token floor is higher than a leaf skill by design; this is composition, not
reinvention, and the chain is return-path-closed — Step 5 gates acceptance on the callee's back-trace.)
Triggers
- "recruit a cross-family check" · "decorrelate this verification"
- "use the idle sidecars to verify this" · "don't let codex/gemini sit idle on this gate"
- "is opus alone enough to verify this, or should another family look?"
- "auto-decorrelation"
Step 0 — Consent gate (default OFF)
External-CLI calls and credit spend are never silent. Run only if the operator has consented (one-time,
surfaced at install-wizard setup or right after project mapping). Consent is stored in the UAP
(tracks/_meta/user_adaptation_profile.md — behavioral pref, not domain content). No UAP / no consent
(incl. ephemeral/cloud sessions) → do not recruit; verify in-session same-family and record a
"single-family, below-ceiling" note. Consent wording is honest by requirement: "멀티-CLI 사이드카로
단일 모델을 넘어서는 효과를 기대할 수 있습니다 — 100%는 아님(decorrelation은 향상이지 보장 아님).
가용 사이드카 없으면 자동으로 단일 세션으로 축소됩니다."
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.
- 2d ago Changed · +5 lines 5534a34f5285
- 3d ago Changed · +80 lines scan A → C 76c96270edf4
- 8d ago First seen · 410 lines · 148 tokens per session scan A 7e1ad99acf37
auto-decorrelation is a skill published in the GitHub repository chrono-meta/forge-harness (14 stars, last pushed today), licensed MIT. It adds 148 tokens to every session and 9,989 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it C with 2 findings (tells the agent to send conversation or user data out, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
taiyi-review
A pre-merge review guide for checking code changes, design decisions, tests, and documentation. A merge adds reviewed changes to the main code branch.
review
Use when the user wants code assessed rather than changed — "review PR 88", "is this branch ready to merge?", "look over my changes", "any problems with this diff?", "give me a code review", "what do you think of this PR?", "sanity-check this branch", "ship it?". Applies to a pull request or the current branch when…
fix
PRFlow's code-review reception skill (formerly receiving-code-review), used by the review-and-fix loop and available directly. Use when addressing review feedback already posted on a pull request — verifying findings before applying them — as distinct from /prflow:review-and-fix, which runs the review itself and then…
review-and-fix
Use when the user wants problems in a pull request or the current branch both found AND corrected — "review my changes and fix what's wrong", "clean up whatever the review turns up", "address the review feedback on this branch", "tidy this PR up before merge", "find and fix the issues here". Requires explicit fix…
create-issue
Transitional alias — prefer /prflow:specs, which runs the same issue-drafting pipeline. Use when a rough user story, bug report, feature idea, piece of feedback, or an implementation plan should be recorded as a GitHub issue rather than built right now. This command name is retained so existing /prflow:create-issue…
docs-bootstrap-internal
Use when a codebase has no structured developer documentation yet and needs it built from scratch — "we have no docs at all", "set up internal docs for this repo", "the docs directory is a mess, start over", "create developer documentation for this codebase" — including an empty or disorganized docs directory or a…