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 instructions/spylogsster/git-ai-review/agents-mdgit clone --depth 1 https://github.com/spylogsster/git-ai-reviewWrote 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/instructions/spylogsster/git-ai-review/agents-md)<a href="https://agentmods.dev/instructions/spylogsster/git-ai-review/agents-md"><img src="https://agentmods.dev/badge/instructions/spylogsster/git-ai-review/agents-md.svg" alt="Measured on agentmods" 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 | $0.01144 | $0.01144 |
| Opus 5 | $0.00572 | $0.00572 |
| Sonnet 5 | $0.00229 | $0.00229 |
| Haiku 4.5 | $0.00114 | $0.00114 |
Grade A, and why
git-ai-review AGENTS.md 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.
How it starts
The opening of the file, as written. The whole thing — 110 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — git-ai-review
This repository contains a publishable npm package for AI-powered Git review hooks. All contributors and AI agents must follow these rules.
- If any rules from other .md files in this repository contradict AGENTS.md, immediately notify the user so the inconsistency can be corrected.
Scope
- This package owns the
git-ai-reviewCLI and related review/hook logic. - Supports Codex CLI, Copilot CLI, and Claude CLI as reviewers (Codex is the primary reviewer).
- The package must remain reusable across arbitrary repositories.
- Avoid repo-specific assumptions except for explicit
AGENTS.md-based policy loading.
Project Structure
- Maintain a tree-based architecture: logically related modules, scripts, files, and resources must be grouped into dedicated directories and subdirectories.
- Do not mix logically unrelated scripts or modules in the same directory.
- When files share a common domain or feature, create a subdirectory for them.
- Merge directories that serve the same logical purpose; split directories that contain unrelated concerns.
Architecture Rules
- Keep logic modular by domain:
- CLI command parsing in
src/cli.ts - review engine in
src/review.ts - pre-commit lock flow in
src/precommit.ts - hook installer in
src/install.ts - prompt/context building in
src/prompt.ts - git utilities in
src/git.ts - review JSON schema in
src/schema.ts - shared types in
src/types.ts
- CLI command parsing in
- Do not collapse domains into a single large file.
- Export testable pure functions for parsing, policy extraction, and verdict logic.
- Never use inline code in npm scripts via
node -eornode --input-type=module -e. Always keep code in dedicated script files. - All imports must be at the top of the file, below the license header section. Inline imports (e.g.,
await import()inside functions) are prohibited. - All changes must work on macOS, Linux, and Windows. Avoid platform-specific shell features (e.g., globstar, POSIX-only commands) in npm scripts; use cross-platform Node.js solutions or well-known npm packages instead.
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.
- 3d ago First seen · 110 lines · 1,144 tokens per session scan A ab7522881123
git-ai-review AGENTS.md is an instructions file published in the GitHub repository spylogsster/git-ai-review (1 stars, last pushed 10d ago), licensed MPL-2.0. It adds 1,144 tokens to every session, about $0.0057 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 instructions, from other repositories
zingit AGENTS.md
Instructions for DanWahlin/zingit, covering zingit - ai agent guide, project overview, architecture, directory structure and tech stack.
atifact AGENTS.md
Instructions for waldekmastykarz/atifact, covering agents and atif spec version.
grok-build-codex-plugin AGENTS.md
Instructions for Reederey87/grok-build-codex-plugin, covering repository guidelines, project structure, using grok in codex, commands and style.
revrem AGENTS.md
AGENTS.md instructions for GitCmurf/revrem, covering agentic coding rules, using meminit, project rules and rules.
strategos AGENTS.md
AGENTS.md instructions for BigBugaboo/strategos, covering strategos contributor guidance, mission, commands, engineering rules and definition of done.
strategos CLAUDE.md
Claude Code instructions for BigBugaboo/strategos: Read and follow AGENTS.md before making changes. Keep edits scoped to the assigned task and do not merge, push, or remove worktrees unless explicitly requested.