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/sioyooo/repogrammar/claude-mdgit clone --depth 1 https://github.com/SioYooo/RepoGrammarWhat 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 | $0.01978 | $0.01978 |
| Opus 5 | $0.00989 | $0.00989 |
| Sonnet 5 | $0.00396 | $0.00396 |
| Haiku 4.5 | $0.00198 | $0.00198 |
Grade A, and why
RepoGrammar CLAUDE.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 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.
This is a copy
100% identical to RepoGrammar AGENTS.md — 0 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 86 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Repository Agent Contract
Authority
- This file and its mirror are the repository's compact, mandatory agent contract. They must remain byte-for-byte identical.
- Read
docs/README.md, then the relevant files under.agents/skills/,.agents/memories/, anddocs/before changing code. - After editing either mirrored guide, immediately run
cargo run --quiet --bin repo-guard -- sync-agent-guides --from <edited-file>. - Do not create nested
AGENTS.md,CLAUDE.md, override files, or competing instruction files without explicit maintainer approval.
Repository Boundaries
- Put all source, executable, test, benchmark, migration-tool, fixture-source, and automation-tool code under
src/. - Outside
src/, only manifests, lockfiles, configuration, documentation, non-code assets, and generated build output are allowed. - Rust tests live beside modules under
#[cfg(test)], insrc/rust/integration_tests/, or in another documented path undersrc/. - Keep nontrivial CI and repository automation logic in
src/rust/bin/repo_guard.rs; declarative workflow files may only invoke it. - Respect the module boundaries and dependency direction defined in
docs/architecture/.
Change Discipline
- Inspect repository status and existing instructions before editing. Preserve unrelated user changes.
- Make the smallest coherent change. Do not perform unrelated refactors, broad formatting, dependency upgrades, or speculative rewrites.
- For nontrivial implementation work, use parallel agent teams where independent slices exist. Assign disjoint ownership, preserve other agents' and users' edits, and integrate results only through the main session after review.
- After implementation, inspect the changed code logic before accepting or merging agent-team output into the main session. Verify behavior with the required checks and resolve conflicts semantically, not by blindly choosing one side.
- During active implementation, batch validation at coherent feature or workstream checkpoints instead of running tests after every small helper or subchange. Run early targeted tests only when they materially unblock diagnosis or protect high-risk behavior; final required checks still gate completion, commit, and merge.
- Every code change must include corresponding tests and documentation changes in the same atomic commit.
- Current v0.1 implementation planning is tracked in
docs/plans/v0.1-parallel-development-plan.md,docs/plans/python-v0.1-implementation-plan.md, and durable memories under.agents/memories/. Update those plan/memory files whenever phase scope, Python v0.1 analysis, CodeGraph provider integration, or UNKNOWN policy changes. - Update normative requirements in
docs/, reusable workflows in.agents/skills/, durable learned context in.agents/memories/, and only cross-cutting mandatory rules in this mirrored contract. - Never leave duplicated requirements inconsistent. Update the canonical document and every affected reference.
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 · 86 lines · 1,978 tokens per session scan A f1a42bff0e58
RepoGrammar CLAUDE.md is an instructions file published in the GitHub repository SioYooo/RepoGrammar (0 stars, last pushed 20d ago), licensed MIT. It adds 1,978 tokens to every session, about $0.0099 per session on Opus 5. A static security scan graded it A with 0 findings. It is 100% identical to RepoGrammar AGENTS.md, differing in 0 lines, and is treated as a copy.
Other instructions, from other repositories
open-design AGENTS.md
AGENTS.md instructions for nexu-io/open-design, covering directory guide, core documentation index, workspace directories, inactive or placeholder directories and development workflow.
ORG2 AGENTS.md
AGENTS.md instructions for org2AI/ORG2, covering agents.md — agent skill routing for orgii, skill routing table, root-cause-first bug fixing, default delivery flow and ui copy conventions.
flameox AGENTS.md
Instructions for morluto/flameox, covering repository guidelines, product direction, project structure & module organization, build, test, and development commands and coding style & naming conventions.
open-design CLAUDE.md
Claude Code instructions for nexu-io/open-design, a project described as: 🎨 Best DeepSeek Harness Design Plugin. The open-source Claude Design alternative. 🖥️ Local-first desktop app. 🖼️ Your coding agent becomes the design engine: prototypes, landing pages, dashboards, slides, images & video — real files…
Agent-Memory-Bridge AGENTS.md
Instructions for zzhang82/Agent-Memory-Bridge, covering agent memory bridge contributor instructions, setup and checks, architecture boundaries, mutation and migration invariants and benchmark expectations.
dreamd AGENTS.md
Instructions for botzrDev/dreamd, covering agents.md — dreamd, what this repo builds, repository layout, build and test and architectural conventions — read before making changes.