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 skills add alexei-led/cc-thingz --skill committing-codegit clone --depth 1 https://github.com/alexei-led/cc-thingzWrote 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/alexei-led/cc-thingz/committing-code)<a href="https://agentmods.dev/skills/alexei-led/cc-thingz/committing-code"><img src="https://agentmods.dev/badge/skills/alexei-led/cc-thingz/committing-code/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/alexei-led/cc-thingz/committing-code"><img src="https://agentmods.dev/badge/skills/alexei-led/cc-thingz/committing-code.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 | $0.00057 | $0.00635 |
| Opus 5.5 | $0.00023 | $0.00254 |
| Sonnet 5.5 | $0.00011 | $0.00127 |
| Haiku 4.5 | $0.00006 | $0.00064 |
Grade A, and why
committing-code 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 15d 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 — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Commit Changes
Group the working-tree changes by purpose and commit them as normal commits. Done
when each intended change is in one logical commit and the final git status is
shown.
Gather state
Run scripts/commit-state.sh gather. It is read-only and prints repo state,
changed paths, diff stats, suspicious paths, and recent commits. Without it, run
git status, git diff --stat, and git log --oneline -8. Read the full diff
(scripts/commit-state.sh full-diff) when the stats do not make the grouping
obvious. Treat the helper output as a hint, not proof.
Stop and report the state verbatim when there is nothing to commit, the directory is not a git repository, HEAD is detached, or a rebase or merge is in progress.
Group
- One purpose, a few files, and no suspicious paths: one commit. Do not invent a split.
- Mixed changes: group by purpose (feature, fix, refactor, docs, CI or config) from diff evidence, not filenames. Keep a change and its tests together.
- Use writing-style for message wording. Keep the repo's syntax and required trailers; lead with the concrete change and add a body only for non-obvious reasons, consequences, or migration steps.
- Match the message style of recent commits. Read conventions.md when there is no history to match or the user asks for a specific format.
Propose and commit
Show the plan and the evidence it rests on:
Based on: git status, diff stat, last 8 commits (style: conventional)
1. fix(validation): reject empty email
- src/validation.ts
- tests/validation.test.ts
2. docs: document the --dry-run flag
- README.md
An explicit commit request authorizes staging and committing this plan. Proceed without a second approval, and do not ask again for each group. Ask first only when the user rejects the grouping, or when the scope, sensitive content, or a history-changing action differs from what was requested.
Hard rules:
- Never stage likely secrets:
.envfiles, private keys, certificates, credential or token files, or files whose content looks like a secret. Flag each one to the user. This check does not replace a secret scanner. - If a pre-commit hook rejects a commit, report its error verbatim. Do not retry
with
--no-verifyand do not amend.
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 15d ago Changed · +4 lines · +13 tokens per session 90d86162adfe
- 22d ago Changed · -4 lines f11da4fe782d
- 1mo ago Changed 0bcb03acd6f0
- 1mo ago First seen · 69 lines · 44 tokens per session scan A 51c32045f19f
committing-code is a skill published in the GitHub repository alexei-led/cc-thingz (38 stars, last pushed 11d ago), licensed MIT. It adds 57 tokens to every session and 635 once invoked, about $0.0002 per session on Opus 5.5. 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
review-delta
Review only changes since last commit using impact analysis. Token-efficient delta review with automatic blast-radius detection.
work-unit-commits
Plan commits as reviewable work units. Trigger: implementation, commit splitting, chained PRs, or keeping tests and docs with code.
gentle-ai-work-unit-commits
Plan commits as reviewable work units. Trigger: implementation, commit splitting, chained PRs, or keeping tests and docs with code.
ring:generating-pr-descriptions
Generating pull request descriptions from git branch changes with automatic title generation, change-type detection, and smart analysis. Uses branch-only scope to avoid full history analysis. Use when preparing a PR for review. Skip when the PR is a single trivial commit or description already exists.
validate-changes
Evaluate staged changes using LLM-as-a-Judge before committing.
merge-contributor-pr
Take a contributor's pull request that cannot be merged as-is — usually a conflict with main — resolve the blocker without rewriting their commits, and merge it with a merge commit so their authorship survives. Detects the case where the conflict exists because the same feature already landed through another PR, and…