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 matheusPavaneli/workbench --skill write-commitgit clone --depth 1 https://github.com/matheusPavaneli/workbenchWrote 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/matheuspavaneli/workbench/write-commit)<a href="https://agentmods.dev/skills/matheuspavaneli/workbench/write-commit"><img src="https://agentmods.dev/badge/skills/matheuspavaneli/workbench/write-commit/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/matheuspavaneli/workbench/write-commit"><img src="https://agentmods.dev/badge/skills/matheuspavaneli/workbench/write-commit.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 | $0.00042 | $0.00637 |
| Opus 5.5 | $0.00017 | $0.00255 |
| Sonnet 5.5 | $0.00008 | $0.00127 |
| Haiku 4.5 | $0.00004 | $0.00064 |
Grade A, and why
write-commit 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 19d 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 — 57 lines — stays where its author put it; the contents beside it link to each section on GitHub.
write-commit
wb <args> # wb not on PATH: python <this skill's directory>/../../lib/wb.py <args>
This skill writes the message. Committing stays the user's call: the CLI prints
the command and runs it only for an explicit --execute.
Steps
-
Learn the house style.
commit convention. It reports what this repo already does — conventional commits, ticket-prefixed subjects, or free-form — from its own history, plus the author identity and whether it matches the context. Follow it; a message in a different style is noise in the log forever. If it reportsfree-formand the team has decided to adopt a style, record it once withcommit convention --set <style>— do not impose one silently. -
See what changed.
git difffor the files,review contextif you need the zones. If a plan exists:sdd get <KEY> --section summarygives the objective in one line. -
Write it.
- Subject: imperative, specific, under 72 characters, no trailing period. "fix expired coupon charging the card" — not "fixes" and not "bug fix".
- Body, when the change is not self-evident: why, not what. The diff already says what. Wrap at 72.
- Reference the ticket key if the repo's history does.
BREAKING CHANGE:in the body if a contract changed.
-
Check it. Write the draft to a file, then
commit check --file <path> --key <KEY>. On pass with a key it is saved to.workflow/<KEY>/commit.txt; on failure nothing is saved and the reasons are listed. Fix the message, not the check. -
Hand it over. Show the message.
git commit <KEY>prints the exact commit command, carrying the saved message and the author the context expects; add--executeonly if the user asked you to commit. It refuses on a protected branch and with nothing staged — choosing what goes in a commit stays theirs.
Rules
- One logical change per commit. If the subject needs an "and", it is probably two commits.
- Never describe what you did not do. No "add tests" in the subject when no test file changed.
- No unfinished markers.
wip,fixup!andsquash!are rejected. - Never put a credential in a message. The check refuses anything that looks like one.
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.
- 19d ago Changed 1025cb880448
- 1mo ago First seen · 57 lines · 42 tokens per session scan A eb173e927511
write-commit is a skill published in the GitHub repository matheusPavaneli/workbench (1 stars, last pushed 19d ago), licensed MIT. It adds 42 tokens to every session and 637 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 2026-08-31.
Other skills, from other repositories
pr-review
Address feedback left on a GitHub pull request: fetch unresolved review threads, make agreed Elixir/Phoenix code fixes, reply, and resolve. Use for a PR URL/number or reviewer comments. NOT for pre-PR review, findings triage, or CI monitoring.
phx-pr-review
Address feedback left on a GitHub pull request: fetch unresolved review threads, make agreed Elixir/Phoenix code fixes, reply, and resolve. Use for a PR URL/number or reviewer comments. NOT for pre-PR review, findings triage, or CI monitoring.
signing-preflight
Before the first commit of a session, check that commit signing will work — the key is loaded and unlocked — and ask the operator to unlock it once, up front, instead of discovering it after a hundred unsigned commits. Never disables signing silently, never re-signs history without asking. Use at the start of…
pr-review
Review pull requests and Git diffs with Graph-It-Live risk, breaking-change, dependency-impact, cycle, unused-export, and test-candidate evidence. Use when asked to review a PR, inspect a diff, assess merge risk, check a branch before merging, create a Graph-It-Live review gate, or explain review-pr results. Trigger…
loop-gh-pr
A recipe for a scheduled GitHub pull-request watcher. It checks open pull requests for new activity and prepares read-only review drafts without posting comments or merging anything.
review
Structured code review with parallel audit agents, confidence-scored triage, and optional auto-fix. Examines uncommitted changes, staged diffs, commit ranges, or specific paths. Produces a tiered report (MUST-FIX / RECOMMENDED / NIT) backed by evidence, then optionally applies fixes with verification.