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/get-tmonier/argot/claude-mdgit clone --depth 1 https://github.com/get-tmonier/argotWrote 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/get-tmonier/argot/claude-md)<a href="https://agentmods.dev/instructions/get-tmonier/argot/claude-md"><img src="https://agentmods.dev/badge/instructions/get-tmonier/argot/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 | $0.04525 | $0.04525 |
| Opus 5 | $0.02263 | $0.02263 |
| Sonnet 5 | $0.00905 | $0.00905 |
| Haiku 4.5 | $0.00453 | $0.00453 |
Grade A, and why
argot 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 5d 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 — 277 lines — stays where its author put it; the contents beside it link to each section on GitHub.
argot
A guardrail that flags code foreign to a repo's own patterns — the dependencies, APIs, and constructs an AI coding agent reaches for that the codebase has never used — learned from the repo's git history (north star + metric: benchmarks/catalogs/RUBRIC.md; novel-pattern catch rate @ low false-alarm). A single statically-linked Rust binary (crates/argot-{core,cli}) — no Python, no Node, no runtime dependencies. (Previously a TS/Bun CLI + Python engine; ported to Rust with verified byte-for-byte parity — see docs/rust-port/.)
Guiding principle
In doubt, optimise for code that's easy to change. The Pragmatic Programmer / craftsmanship lens: the right design is the one a future contributor (human or agent) can extend, refactor, or revert without archaeology. When two options look equally correct, pick the one with the smaller blast radius and clearer seams. Don't add abstractions before the second use case shows up; don't keep dead code "just in case"; don't suppress a check when the underlying code is the real fix. Strict tooling (clippy -D warnings, the golden suites) exists to surface change-cost early — work with it, not around it.
Strategy & positioning
Before changing Argot's product, positioning, website, or public claims, read:
FOUNDER.md— the one-page operating manifesto.docs/strategy/ARGOT_STRATEGY.md— canonical strategy (normative decision register D1–D14).docs/strategy/ARGOT_CURRENT_REALITY.md— authoritative on what ships today; never market a product requirement as current reality.
This is a founder-level operating layer; it does not change the engineering conventions below.
Task runner
Always use just — it's the canonical interface for all dev commands.
just verify # fmt --check + clippy -D warnings + test, then verify-features
just verify-features # the same, per feature: script · arch · integrity
just test # cargo test --workspace
just extract . # run extract on this repo → .argot/dataset.jsonl
just dogfood # run full pipeline against argot itself (or any path) — fast monorepo check
just build # cargo build --release -p argot → target/release/argot
just bench-quick # ~1 min bench smoke (one fixture per category + 50 controls)
just arch-verify # ~25 s architecture-layer fixture-recall regression guard
just integrity-verify # gaming-fixture recall + control guard for the integrity rules
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.
- 5d ago First seen · 277 lines · 4,525 tokens per session scan A a534abbc9554
argot CLAUDE.md is an instructions file published in the GitHub repository get-tmonier/argot (48 stars, last pushed 3d ago), licensed MIT. It adds 4,525 tokens to every session, about $0.0226 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 2026-08-30.
Other instructions, from other repositories
habit-hooks CLAUDE.md
Instructions for habit-hooks/habit-hooks, covering habit-hooks notes, architecture, gotchas, a tool that resolves symlinks now hard-fails its sensor and jsdoc nodes are not multilinecommenttrivia in ts-morph.
compose-doctor CLAUDE.md
Instructions for rotemmiz/compose-doctor, covering compose-doctor, status, architecture, scoring and conventions.
compose-doctor AGENTS.md
Instructions for rotemmiz/compose-doctor, covering agents.md — compose-doctor, what compose-doctor is, the fix loop, integrity (non-negotiable) and writing good compose in the first place.
invigil AGENTS.md
AGENTS.md instructions for invigil/invigil, covering agents.md — conventions for ai coding agents, build / test / lint and layout.
anma CLAUDE.md
Instructions for anma-labs/anma, covering anma, commands, internal architecture, what not to edit by hand (generated by anma sync) and dogfooding note (known limitation).
membrowse-action CLAUDE.md
Claude Code instructions for membrowse/membrowse-action, covering claude.md, project overview, command-line interface, unified cli (membrowse) and output as json instead.