Borrowing it
Nothing to install: this file belongs to FilipNowakowicz/nixos-config. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/FilipNowakowicz/nixos-config/main/.agents/skills/review-learning-candidates/SKILL.mdgit clone --depth 1 https://github.com/FilipNowakowicz/nixos-configWrote 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/filipnowakowicz/nixos-config/review-learning-candidates)<a href="https://agentmods.dev/skills/filipnowakowicz/nixos-config/review-learning-candidates"><img src="https://agentmods.dev/badge/skills/filipnowakowicz/nixos-config/review-learning-candidates/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/filipnowakowicz/nixos-config/review-learning-candidates"><img src="https://agentmods.dev/badge/skills/filipnowakowicz/nixos-config/review-learning-candidates.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00039 | $0.01617 |
| Opus 5 | $0.00019 | $0.00809 |
| Sonnet 5 | $0.00008 | $0.00323 |
| Haiku 4.5 | $0.00004 | $0.00162 |
Grade A, and why
review-learning-candidates 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 11d 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 — 138 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Review Learning Candidates
Review compact learning candidates under .agents/learning/candidates/.
Candidate review is explicit, batch-oriented work. Do not run it as part of
normal task wrap-up.
End state of a review pass: every candidate touched is either implemented in a
PR (and promoted), superseded, rejected, or consciously left open with a
reason. The promotion and the candidate's status flip travel together in the
same branch, and promoted/rejected/superseded files are moved out of the
open queue so the index stays lean.
Goal
Turn candidate proposals into the strongest useful repo artifact:
- assertion / test / CI gate
- hook
- skill
- doc
- rejection
Prefer executable enforcement over prose. A candidate that can become a check,
test, or invariant should not be promoted to CLAUDE.md.
For candidates already promoted by review and routed implement-fix with
best_form: test, best_form: assertion, best_form: invariant, or
best_form: check, use
.agents/learning/scripts/scaffold-promotion-check.sh to create the failing
check stub and scripts/validate.sh wiring before writing prose. The generated
stub is intentionally red until you replace its TODO body with the real
assertion, and it keeps a back-reference to the candidate for audit.
Workflow
1. Triage from metadata
- Run
bash .agents/learning/scripts/validate-candidates.sh. - Run
bash .agents/learning/scripts/review-candidates.sh— read status counts, the route/form grouping, and the Expired open candidates list. - Run
bash .agents/learning/scripts/check-candidate-liveness.sh— it flags open candidates whose evidence already landed (PR merged / commit on HEAD). - Choose a small batch by
route,best_form, or relatedtargets. Open only the candidate files in that batch.
2. Decide one outcome per candidate
Resolve liveness first: if check-candidate-liveness.sh reports LIKELY
RESOLVED, or you confirm by hand that the fix already landed, the route is
moot — mark it superseded and do not re-implement. NO SIGNAL is not a clean
bill of health (the evidence just lacks a pointer); judge it from evidence and
targets. Treat an implement-fix candidate as "is a fix still needed?", never
as "a fix is owed" — the fix is often already in tree.
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.
- 11d ago First seen · 138 lines · 39 tokens per session scan A 39749f305328
review-learning-candidates is a skill published in the GitHub repository FilipNowakowicz/nixos-config (5 stars, last pushed yesterday), licensed MIT. It adds 39 tokens to every session and 1,617 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 2026-08-31.
Other skills, from other repositories
quality-code
Standards for writing or modifying handwritten source code. Use when implementing code changes or reviewing handwritten code. Do not use for browsing, explanation, diagnosis without implementation, generated code, or vendored code.
code-qa
Review handwritten source code against the quality-code standards. Use only when the user explicitly invokes $code-qa. Do not infer use from an ordinary request to review, inspect, or explain code.
autoreview
Pre-commit/ship code review: Codex default; optional Claude or Pi.
omh-code-review
This is a Hermes-native code-review workflow skill.
code-reviewer
Code review specialist focused on patterns, bugs, security, and performance.
full-repo-review
Comprehensive four-wave review of all repo source files, producing a prioritized issue backlog.