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 agentmods add skills/thoughtbot/rails-consultant/git-commitnpx skills add thoughtbot/rails-consultant --skill git-commitgit clone --depth 1 https://github.com/thoughtbot/rails-consultantWhat 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 | $0.00047 | $0.02126 |
| Opus 5 | $0.00023 | $0.01063 |
| Sonnet 5 | $0.00009 | $0.00425 |
| Haiku 4.5 | $0.00005 | $0.00213 |
Grade A, and why
git-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 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.
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 — 157 lines — stays where its author put it; the contents beside it link to each section on GitHub.
What this does
Look at everything that has changed in the working tree, decide how the changes should be split into commits, and then create those commits — messages and all — without stopping to ask for approval. The user invoked /git-commit because they trust you to make the call, so make it and report what you did afterward.
Any argument the user passes is context, not a command: a hint about intent ("addressing review feedback", "this is risky, note the migration") that should inform how you group and what you write. It is never the literal commit message.
Step 1: Understand what changed
Before grouping anything, build a real picture of the diff. Run these together:
git status— what's modified, added, deleted, untrackedgit diff— unstaged changesgit diff --staged— anything already stagedgit log --oneline -15— recent history, to match the tone and see referenced issues/PRs
Read the diff to understand the why, not just the what. You are about to explain these changes to a future reader; you can't do that if you only know which lines moved. If something is genuinely unclear, it's fine to ask one focused question — but usually the diff plus recent history tells the story.
If there's nothing to commit, say so and stop. Never create an empty commit.
Step 2: Decide the grouping — this is the judgment call
The goal is atomic commits: each commit is one complete, coherent change that stands on its own. The hard part is knowing how many commits that means, and there's no mechanical rule — it depends on what the work actually is.
Lean toward a single commit when the changes tell one story. If the user just built a feature — new model, its migration, the controller, the view, the tests — that's one logical change even though it spans many files. Splitting it into "add migration", "add model", "add tests" produces commits that are individually broken and useless to revert. Keep it together.
Split into separate commits when the changes are genuinely independent. The tell is that the work is a set of unrelated things rather than one thing:
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 First seen · 157 lines · 47 tokens per session scan A 6dcaa130dd70
git-commit is a skill published in the GitHub repository thoughtbot/rails-consultant (22 stars, last pushed 1mo ago), licensed MIT. It adds 47 tokens to every session and 2,126 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 2026-08-30.
Other skills, from other repositories
upgrade-cleanup
Clean up after (or abandon) a Rails upgrade. Drop NextRails.next? and NextRails.current? branches and retire dual-boot scaffolding (Gemfile.next, Gemfile.next.lock, conditional Gemfile groups), keeping either the next or the current version. Trigger when the user says they are done with the upgrade, want to clean up…
recursive-decomposition
Handle tasks that exceed the context window by decomposing them: size and filter the input, chunk it, run recursive sub-agents on independent parts, verify on small windows, and synthesise programmatically, following the Recursive Language Models (RLM) research by Zhang, Kraska and Khattab (2025). Use when a task…
typescript-react-nextjs-patterns
Production-grade TypeScript reference for React & Next.js frontend development. Covers type narrowing, component Props, generic hooks, discriminated unions, as const, satisfies, Zod validation, TanStack Query, server/client boundaries, forms, state management, performance, accessibility, debugging, and code review.…
swarm
Run a multi-agent audit of a codebase by spawning specialized parallel subagents (security, performance, tests, architecture, dead-code), then synthesize their findings into a single prioritized action plan. Use this whenever the user runs /swarm, asks to "audit the repo," "review this codebase," "find issues across…
crisp-execute
CRISP Execute — Sprint execution loop with change request management, security gates, product gates, and stakeholder reporting. Use after Phase S (Spec) is complete and crisp-state.json shows readyforexecute: true. Triggers on "start sprint", "begin build", "execute", "sprint 1", "run sprints", or when handed off from…
phase1-clarify
The Mileva Method (CRISP) — Phase 1: Clarify. Problem definition, eliciting, painkiller test, buy vs build, AI justification gate, Go/No-Go. Use at the start of any AI implementation project. Triggers on "clarify", "phase 1", "start discovery", "define the problem", "what problem are we solving", or at the beginning…