git-manager

git-manager is an agent for Claude Code from duc01226/EasyPlatform. It costs 37 tokens per session (6,593 once invoked), scanned A, original, MIT.

A Git workflow agent that stages changes, checks them for secrets, creates conventional commits, and pushes them when explicitly requested.

In plain words
What is it for?
Use it to prepare commits after a feature or fix, split mixed changes into logical commits, and publish them to a remote repository.
Why use it?
It helps keep commit history organized and prevents accidental commits containing sensitive information.

Agent for Claude Code

Written for Claude Code: SessionEnd hook event. Also seen: model in frontmatter; reads .claude/ paths; mentions subagents.

Needs its repository: it runs a file that does not travel with it, so clone the repository first. The line is | node .claude/hooks/lib/git-operation-lease.cjs issue.

Good fit Use it to prepare commits after a feature or fix, split mixed changes into logical commits, and publish them to a remote repository.

Compare 6 agents from other repositories ↓
Install

Getting it into your agent

It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.

Clone the repo
git clone --depth 1 https://github.com/duc01226/EasyPlatform
agentmods
npx agentmods add agents/duc01226/easyplatform/git-manager

Made for: Claude Code.

Wrote 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.

agentmods badge for git-manager

README.md
[![agentmods](https://agentmods.dev/badge/agents/duc01226/easyplatform/git-manager/github.svg)](https://agentmods.dev/agents/duc01226/easyplatform/git-manager)
Your own site
<a href="https://agentmods.dev/agents/duc01226/easyplatform/git-manager"><img src="https://agentmods.dev/badge/agents/duc01226/easyplatform/git-manager/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.

agentmods 80×15 button for git-manager

Your own site · 80×15
<a href="https://agentmods.dev/agents/duc01226/easyplatform/git-manager"><img src="https://agentmods.dev/badge/agents/duc01226/easyplatform/git-manager.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 37 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 6,593 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00037 $0.06593
Opus 5.5 $0.00015 $0.02637
Sonnet 5.5 $0.00007 $0.01319
Haiku 4.5 $0.00004 $0.00659

Measured 3d ago against content hash 56b023a304b0, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

git-manager 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 3d 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.

.claude/agents/git-manager.md · 441 lines

How it starts

The opening of the file, as written. The whole thing — 441 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Connected Skill Contracts

Skill connection: Apply the task-specific procedure from the connected canonical skill contract that matches the assigned brief. The role-specific quality SYNC blocks in this prompt are the static sub-agent quality protocol; do not expand orchestrator-only instructions inside a leaf assignment.

Connected contracts:

  • commit

Quick Summary

Goal: Perform only the explicitly requested Git operation and scope, targeting 2-4 tool calls where practical. For commits, produce secret-free, conventional-commit history split into logical commits when types/scopes mix.

Summary: (read-this-if-nothing-else digest — purpose + ALL main steps + gates)

  • AUTHORITY FIRST — validate operation, scope, and sourceRequest against the actual explicit user request using the Git Request Contract below. Implementation completion and review approval never authorize Git.
  • PURPOSE — execute only that operation; for commits, create secret-free conventional-commit history, splitting logically when types/scopes mix. Push requires explicit user authority.
  • STEP 1 — AUTHORITY + STAGE + SCAN. Inspect the current index first, issue a short lease only for the user's literal operation(s), then stage only authorized files, count lines/files, scan for secrets, and classify file groups — read that output ONCE. A stage-only request stops after staging/scanning; a push-only request skips staging and commit preparation but still requires a push lease.
  • GATE A (hard) — SECRETS > 0 → STOP and block. Report file/line/rule/category and redacted metadata only; NEVER commit through it.
  • STEP 2 — SPLIT DECISION. Types/scopes mixed (feat+fix, code+deps, config+features) → multiple commits; small single-scope change → one.
  • STEP 3 — GENERATE MESSAGE(S). Simple: author directly. Complex (LINES > 30 OR FILES > 3): gemini CLI, falling back to authoring it yourself if unavailable.
  • STEP 4 — COMMIT via printf … | git commit -F -, then push ONLY if the user asked. Revoke every lease in a finally path, including failed hooks or failed Git commands.
  • GATE B (hard) — "push" happens ONLY when the user literally said push, and NEVER directly to main/master — those land via PR.
  • OUTPUT — terse results only (<1k chars), no narration of what you did.

Read the full file on GitHub · 441 lines

Changes

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.

  1. 3d ago Changed · -184 lines · +5 tokens per session 56b023a304b0
  2. 1mo ago First seen · 625 lines · 32 tokens per session scan A 3ba914afb3d4

Subscribe to this mod's changes

git-manager is an agent published in the GitHub repository duc01226/EasyPlatform (12 stars, last pushed yesterday), licensed MIT. It adds 37 tokens to every session and 6,593 once invoked, about $0.0001 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 2026-09-03.