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/isaiascope/ai/iso-commitnpx skills add IsaiaScope/ai --skill iso-commitgit clone --depth 1 https://github.com/IsaiaScope/aiWhat 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 | $0.00103 | $0.01996 |
| Opus 5 | $0.00051 | $0.00998 |
| Sonnet 5 | $0.00021 | $0.00399 |
| Haiku 4.5 | $0.00010 | $0.00200 |
Grade A, and why
iso-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 yesterday.
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.
iso-commit
Turn the current changes into one commit with a message a human will still understand in six months.
Invocation: /iso-commit [--staged] [--split].
| flag | effect |
|---|---|
| (none) | git add -A — staged, unstaged, and untracked — then one commit |
--staged |
commit the index only; stage nothing |
--split |
group the diff by concern and make several commits instead of one |
Never commit unless the user asked. /iso-commit typed by the user is the only trigger. Do not commit at the end of iso-write, after making edits, or because a task "feels done" — the rest of the iso-* chain deliberately leaves the tree uncommitted, and this skill does not change that.
Flow
Mechanics live in skills/iso-commit/scripts/commit.sh. Run it by absolute path.
- Preflight —
commit.sh preflight [--staged]. Non-zero: print its message and stop. Checks repo, branch (rejects detached HEAD), and that there's something to commit. - Read the change —
git diff HEAD(orgit diff --cachedwith--staged), plusgit log --oneline -10to match the repo's existing subject style. - Stage —
commit.sh stage [--staged]. Runs the credential guard first and exits 2 without touching the index if it trips. On exit 2, print the blocked paths and stop; do not work around it. - Write the message to a temp file, following the format below.
- Branch gate —
commit.sh gate "<subject>", thencommit.sh land <action> <branch>. See below. - Commit —
commit.sh commit <msgfile>. Hooks run on purpose (commitlint gates the subject, version-bump reads it). If a hook rejects the commit, fix the message and retry — never--no-verify. - Report — print the message and the short SHA the script echoed. Add one line:
wrong? git commit --amend.
Step 5 — the branch gate
dev, test and prod are promoted into, never worked on: a commit
made while standing on one is stranded, invisible to the cascade, and reachable
only by a git reset --hard typed by hand. The gate catches that before the
commit exists, rather than after.
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.
- yesterday First seen · 157 lines · 0 tokens per session scan A daa08fdbe73d
iso-commit is a skill published in the GitHub repository IsaiaScope/ai (2 stars, last pushed 4d ago), licensed MIT. It adds 103 tokens to every session and 1,996 once invoked, about $0.0005 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
agent-architecture-planner
Use when designing an autonomous agent, planning agent architecture, building a scheduled automation, or creating a Claude Code agent workflow. Triggers: 'design an agent', 'build an automation', 'agent architecture', 'automate this workflow', 'create a scheduled agent', 'shell script agent'.
81-pipecat
Create your Pipecat skill from official documentation, then learn to improve it throughout the chapter.
agent-audit
Validates agent configurations for model selection, tool permissions, focus areas, and approach quality. Use when reviewing, auditing, improving agents, or learning agent best practices.
2026-legal-research-agent
Expert legal research agent for finding and scraping expungement data state by state. Knows authoritative sources, URL patterns, Firecrawl configuration, and 2026 legal landscape. Activate on "find expungement data", "scrape state laws", "legal research", "court URLs", "statute sources", "Clean Slate laws", "automatic…
80-livekit-agents-majiayu000-claude-skill-registr
Create your LiveKit Agents skill from official documentation, then learn to improve it throughout the chapter.
ac-lock
Phase 4.0 — Acceptance Criteria lock checkpoint before implementation. Consolidates all ACs from PRD, TechSpec, and Tasks, presents to user for confirmation, and saves the locked AC list to accepted-criteria.md. Implementation only starts after explicit user confirmation.