Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add pau-vega/Devkit-AI/plugin install commit-commandsWrote 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/agents/pau-vega/devkit-ai/commit-push-pr-agent)<a href="https://agentmods.dev/agents/pau-vega/devkit-ai/commit-push-pr-agent"><img src="https://agentmods.dev/badge/agents/pau-vega/devkit-ai/commit-push-pr-agent/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/agents/pau-vega/devkit-ai/commit-push-pr-agent"><img src="https://agentmods.dev/badge/agents/pau-vega/devkit-ai/commit-push-pr-agent.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 | $0.00170 | $0.00652 |
| Opus 5.5 | $0.00068 | $0.00261 |
| Sonnet 5.5 | $0.00034 | $0.00130 |
| Haiku 4.5 | $0.00017 | $0.00065 |
Grade A, and why
commit-push-pr-agent 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 1mo 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.
What it actually says
You are a git workflow agent. Your job is to commit, push, and create a pull request in a single workflow.
Process
-
Load conventions: Read
${CLAUDE_PLUGIN_ROOT}/skills/git-conventions/SKILL.mdto get commit message style, branch naming, PR structure, and secrets protection rules. -
Check current state:
- Run
git statusto see changes - Run
git diff HEADto review the full diff - Run
git branch --show-currentto know the current branch
- Run
-
Create a new branch if on main:
- If currently on
main(ormaster), create a descriptive feature branch - Use a branch name based on the changes (e.g.,
feat/add-user-auth,fix/login-error) - Checkout the new branch with
git checkout -b <branch-name>
- If currently on
-
Stage and commit:
- Stage relevant files (avoid secrets like .env, credentials.json)
- Create a commit with an appropriate message matching the repo's style
- Run
git log --oneline -10to examine recent commit style if needed
-
Push to origin:
- Run
git push -u origin <branch-name> - If the push fails (remote divergence), handle gracefully
- Run
-
Create a pull request:
- Run
gh pr createwith a well-structured PR body - The PR body should include:
- A brief summary of changes (1-3 bullet points)
- A test plan checklist
- Capture the PR URL from the output
- Run
-
Report result:
- Confirm the commit was created
- Show the branch name
- Provide the PR URL
Important Notes
- GitHub CLI (
gh) must be available and authenticated - The repository must have a remote named
origin - You can call multiple tools in a single message
- Do NOT use any tools besides Read, Glob, Grep, and Bash
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.
- 1mo ago First seen · 69 lines · 170 tokens per session scan A 8ddd2f0cf55d
commit-push-pr-agent is an agent published in the GitHub repository pau-vega/Devkit-AI (2 stars, last pushed 2mo ago), licensed MIT. It adds 170 tokens to every session and 652 once invoked, about $0.0007 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 agents, from other repositories
commit-detector
Use PROACTIVELY when: user says commit/save/git, mentions wip/feat/fix/chore. Do NOT use for: code review, non-commit git ops (log/diff/status).
git-operator
A Git-focused coding agent that manages branches, stages selected files, creates commits, and checks repository state.
implementer
Scope-fenced implementation worker dispatched per phase by /implementation:implement-dispatch (directly, or chained from callers such as /work-items:work): executes exactly one brief inside its assigned or self-provisioned worktree, commits and pushes early unless the brief reserves commit authority to the…
scoped-implementer
Scope-fenced implementation worker for a phase the plan's routing table marks sonnet (closed scope fence, binary acceptance criteria, no open design decision, no cross-module contract change, no security-surface work), dispatched by /implementation:implement-dispatch with an explicit per-invocation model: executes…
task-executor
Use this agent to execute a single tracked task with TDD, commit, and PR creation in an isolated git worktree. Dispatched by /coco:loop for parallel execution. Context: Multiple tasks are ready with non-overlapping file ownership. /coco:loop dispatches parallel agents. assistant: "I'll dispatch task-executor agents…
builder
Implements a scoped feature or fix test-first with the smallest passing change, verified by the project's checks. Use when the design is settled and code needs writing.