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 AkshitIreddy/agent-skills --skill atomic-commitsgit clone --depth 1 https://github.com/AkshitIreddy/agent-skillsWrote 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/akshitireddy/agent-skills/atomic-commits)<a href="https://agentmods.dev/skills/akshitireddy/agent-skills/atomic-commits"><img src="https://agentmods.dev/badge/skills/akshitireddy/agent-skills/atomic-commits/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/akshitireddy/agent-skills/atomic-commits"><img src="https://agentmods.dev/badge/skills/akshitireddy/agent-skills/atomic-commits.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 | $0.00045 | $0.00533 |
| Opus 5 | $0.00023 | $0.00267 |
| Sonnet 5 | $0.00009 | $0.00107 |
| Haiku 4.5 | $0.00005 | $0.00053 |
Grade A, and why
atomic-commits 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 9d 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 — 40 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Atomic commits with Conventional Commits
Rules for every commit in every repository, regardless of project or language.
Scope: one concern per commit
- Each commit contains exactly one logical change: one feature, one fix, one refactor, one doc update. If the summary line needs "and", split the commit.
- Never bundle unrelated changes (e.g. a bug fix plus a rename plus a dependency bump) — stage selectively (
git add -p, per-file staging) so each commit stays scoped. - Prefer many small commits over one large one, but each commit must stand on its own: it should make sense in
git logwithout the neighbors.
Only commit working states
- Every commit should build/parse and leave the project in a coherent, non-broken state. Run the relevant build/tests before committing when practical.
- Don't commit half-finished work as a checkpoint on shared branches. If a change can't work alone, fold it into the commit that completes it.
- Never commit generated artifacts, secrets, or local-only config; keep
.gitignorecurrent as part of the change that introduces new artifact types.
Message format: Conventional Commits
<type>(<optional scope>): <imperative summary, lowercase, no trailing period>
<optional body: what and why, wrapped ~72 cols>
- Types:
feat,fix,refactor,perf,docs,test,build,ci,chore,style,revert. - Scope is the area touched (
feat(engine): …,fix(ui): …) — use it when the repo has distinct areas. - Summary ≤ 72 chars, imperative mood ("add", not "added"/"adds"), specific:
fix(parser): handle empty CSV cellsnotfix stuff. - Body only when the diff doesn't explain itself — state the why, constraints, or tradeoffs, not a restatement of the diff.
- Breaking changes:
!after type/scope (feat(api)!: …) plus aBREAKING CHANGE:footer describing the migration.
Ordering
- Commit infrastructure/scaffolding before the code that depends on it, so history replays cleanly.
- Refactors that enable a feature go in their own
refactor:commit before thefeat:commit.
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.
- 9d ago First seen · 40 lines · 45 tokens per session scan A 365617d43099
atomic-commits is a skill published in the GitHub repository AkshitIreddy/agent-skills (1 stars, last pushed 12d ago), licensed MIT. It adds 45 tokens to every session and 533 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
commit-staged
This skill should be used when user asks to "commit these changes", "write commit message", "stage and commit", "create a commit", "commit staged files", or explicitly invokes "commit-staged".
lov-checkpoint-list
A read-only history report built from Git records and project checkpoint logs. A checkpoint is a recorded project milestone or state, while Git is the version-control system that tracks changes and commits.
lov-git-commit-with-context
A Git commit helper that compares the current conversation's intended work with the repository's actual changes before creating a commit. Git is a system for tracking file changes, and a commit records a selected set of those changes.
chinese-commit-conventions
A Chinese-language convention for writing Git commit messages, based on Conventional Commits, a standard format that tools can use to classify changes and build changelogs.
signal-commit
Stage all changes and create a conventional commit from the diff. Zero prompts, zero confirmation by default. Use when user types /signal-commit, "commit everything", "just commit", "signal commit", or asks to commit current changes without specifying a message. Supports --draft (show message without committing) and…
signal-pr
Stage all changes, commit, push, then open a pull request. Title and body are generated from the diff and commit history. Use when user types /signal-pr, "open a PR", "create pull request", "ship a PR", or "make a PR". Requires gh CLI.