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/dpeluche/trs/claude-mdgit clone --depth 1 https://github.com/dPeluChe/trsWrote 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/dpeluche/trs/claude-md)<a href="https://agentmods.dev/instructions/dpeluche/trs/claude-md"><img src="https://agentmods.dev/badge/instructions/dpeluche/trs/claude-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 | $0.00529 | $0.00529 |
| Opus 5 | $0.00264 | $0.00264 |
| Sonnet 5 | $0.00106 | $0.00106 |
| Haiku 4.5 | $0.00053 | $0.00053 |
Grade A, and why
trs 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 4d 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 — 47 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
Agent instructions for this repo live in AGENTS.md. Start
there. This file carries only the machine-readable config the hooks and the
ship gate read.
Personal agent setup (skills, flow imports) belongs in your own global config, not here: this file is committed and shared by every contributor.
ship config
# Mirrors .github/workflows/ci.yml so a local gate fails whatever CI would.
# --all-targets is a superset of CI's plain clippy: it also lints tests.
lint: cargo fmt -- --check && cargo clippy --all-targets -- -D warnings
# No separate typecheck: clippy type-checks the crate, so a `cargo check`
# step would just repeat the same work.
build: cargo build
# --no-fail-fast so one failing suite doesn't mask the rest (matters most
# on the blocking windows-latest job).
test: cargo test --no-fail-fast
# pre-push deliberately not delegated: it already mirrors CI and is stricter.
hooks_skip: pre-push: "already mirrors CI (fmt + clippy + full suite), stricter than the ship-config gate"
merge_policy: ask # auto | ask
loc_limit: 500
simplify: 500 # run /simplify only if changed LOC > N (off = only on request)
Git hooks
core.hooksPath points at the tracked .githooks/, so lefthook's own stubs
in .git/hooks/ are never invoked. .githooks/pre-commit and
.githooks/commit-msg delegate to lefthook explicitly; both no-op when
lefthook isn't installed, so nobody is blocked by an optional tool.
.githooks/pre-push deliberately does NOT delegate: it already mirrors CI
(fmt + clippy + the full test suite), which is stricter than the
ship-config gate, and delegating would run clippy twice.
Release
Tagging vX.Y.Z triggers .github/workflows/release.yml (5 binaries + npm
publish + GitHub Release). Bump Cargo.toml, regenerate CHANGELOG.md with
git cliff -o CHANGELOG.md --tag vX.Y.Z, merge, then tag. A tagged run uses
the workflow at the tag's commit, so fixing a failed release means re-tagging,
not re-running.
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.
- 4d ago First seen · 47 lines · 529 tokens per session scan A bb29e64304f6
trs CLAUDE.md is an instructions file published in the GitHub repository dPeluChe/trs (10 stars, last pushed 6d ago), licensed MIT. It adds 529 tokens to every session, about $0.0026 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
obsidian-mcp-rs CLAUDE.md
Instructions for MrRefactoring/obsidian-mcp-rs, covering claude.md, commands, build / run, tests and lint gates enforced by ci (.github/workflows/ci.yml).
codemark GEMINI.md
Gemini CLI instructions for DanielCardonaRojas/codemark, covering development workflow and debugging with the tui logging system.
hiveshare CLAUDE.md
Instructions for KB-perByte/hiveshare, covering hiveshare — claude.md, repo layout, build & run, key env vars (server) and naming: hive vs memory.
llm-safe-haven CLAUDE.md
Claude Code instructions for pleasedodisturb/llm-safe-haven, covering llm safe haven, what this is, project structure, tdd — non-negotiable (adopted 2026-08-17) and the contract.
bullet-bots AGENTS.md
Instructions for bulletxyz/bullet-bots, covering bullet-bots, build & test, run the faucet curl printed by keygen, then, or (base58 from phantom/delegation export, or hex) and architecture — the harness, feeds, and actors.
kleosrules AGENTS.md
Instructions for kleosr/kleosrules, a project described as: Cursor harness pack: user rules, skills, Bash hooks, local HANDOFF memory. macOS, Linux, Windows (WSL).