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/codeready-toolchain/tarsy/agents-mdgit clone --depth 1 https://github.com/codeready-toolchain/tarsyWrote 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/codeready-toolchain/tarsy/agents-md)<a href="https://agentmods.dev/instructions/codeready-toolchain/tarsy/agents-md"><img src="https://agentmods.dev/badge/instructions/codeready-toolchain/tarsy/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 | $0.00550 | $0.00550 |
| Opus 5 | $0.00275 | $0.00275 |
| Sonnet 5 | $0.00110 | $0.00110 |
| Haiku 4.5 | $0.00055 | $0.00055 |
Grade A, and why
tarsy 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 — 58 lines — stays where its author put it; the contents beside it link to each section on GitHub.
TARSy - agent guide
TARSy is a hybrid Go + Python SRE system: Go orchestrator (alerts, sessions, MCP tools), a stateless Python LLM service over gRPC, and a React dashboard.
Default branch: master. Architecture and package map: CLAUDE.md.
Commands
Never invent alternate package-manager commands if these work.
- Full local gate:
make check-all - Lint:
make lint(make lint-fixto auto-fix) - Test:
make test - Format:
make fmt - Dev:
make doctor,make setup,make dev,make dev-stop - Go:
make test-unit,make test-go,make build - Python LLM service:
make test-llm - Dashboard:
make test-dashboard - Codegen:
make ent-generate,make proto-generate - Migrations:
make migrate-create NAME=add_feature— then apply thedb-migration-reviewskill - LLM image deps: after changing
llm-service/uv.lock, runmake llm-requirements
Skills
Project skills live in .cursor/skills/ (symlinked as .claude/skills/). Load a skill when its description matches the task. Do not preload all of them.
How we work
- Don't assume; surface tradeoffs and ask when unclear
- Minimum code that solves the problem; no speculative features or abstractions
- Touch only what you must; no drive-by refactors
- Define success criteria and verify before finishing
Commits and PRs
- Never commit, amend, or push unless the user explicitly asks in this turn
- “Make the change”, “fix the tests”, or finishing a task is not permission to commit
- If it is unclear, leave changes uncommitted and ask
- Conventional commits:
feat|fix|docs|chore|refactor|test(scope): summary - Prefer small, focused PRs
- Fixes should reference an issue:
Fixes #123
Style
- Match existing code; do not reformat unrelated files
- Prefer clear names over clever abstractions
- No AI walls of text in PR descriptions
Safety
- Do not commit secrets
- Do not weaken CI, auth, or permission checks without an explicit human request
- Ask before large architectural 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.
- 3d ago First seen · 58 lines · 550 tokens per session scan A 06c3072c46b6
tarsy AGENTS.md is an instructions file published in the GitHub repository codeready-toolchain/tarsy (10 stars, last pushed 4d ago), licensed Apache-2.0. It adds 550 tokens to every session, about $0.0028 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
siclaw CLAUDE.md
Instructions for scitix/siclaw, covering siclaw — operating manual for claude, what this project is, critical architecture invariants, 🔴 local mode: shared filesystem and 🔴 skill bundle contract.
siclaw AGENTS.md
Instructions for scitix/siclaw, covering siclaw project instructions, project shape, before risky edits, verification and stable boundaries.
rootly-graphify-importer AGENTS.md
Instructions for Rootly-AI-Labs/rootly-graphify-importer, covering rootly-graphify, graphify pipeline — full instructions, usage patterns, step 1 — ensure graphify is installed and step 2 — detect files.
runwhen-platform-mcp AGENTS.md
AGENTS.md instructions for runwhen-contrib/runwhen-platform-mcp, covering agents.md — runwhen platform mcp server, authentication, important: workspacename is required, tool routing — workspacechat vs direct tools and always prefer workspacechat for.
runwhen-platform-mcp copilot-instructions.md
Copilot instructions for runwhen-contrib/runwhen-platform-mcp, covering runwhen platform mcp server — copilot instructions, workspacename is required, tool routing, available skills and script contract (quick reference).
ai-reliability-copilot AGENTS.md
AGENTS.md instructions for YanpengQi7/ai-reliability-copilot: This version has breaking changes — APIs, conventions, and file structure may all differ from your training data. Read the relevant guide in nodemodules/next/dist/docs/ before writing any code. Heed deprecation notices.