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/tjboudreaux/rosetta/agents-mdgit clone --depth 1 https://github.com/tjboudreaux/rosettaWrote 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/tjboudreaux/rosetta/agents-md)<a href="https://agentmods.dev/instructions/tjboudreaux/rosetta/agents-md"><img src="https://agentmods.dev/badge/instructions/tjboudreaux/rosetta/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 | $0.01261 | $0.01261 |
| Opus 5 | $0.00630 | $0.00630 |
| Sonnet 5 | $0.00252 | $0.00252 |
| Haiku 4.5 | $0.00126 | $0.00126 |
Grade A, and why
rosetta 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 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 — 84 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Briefing for AI agents working in this repository. Keep it high-signal.
Rosetta is an open-source agent skill that reconciles every coding agent's local conversation
history (18 transcript stores) with git history and docs into one cited ground truth, plus
durable decision records (ADR / PDR / BDR). The shipped collector and decision CLI are
pure-stdlib Python with zero runtime dependencies (ADR 0013). The skill itself lives under
skills/rosetta/; the repo root holds the README, llms.txt, assets/, and LICENSE.
Build & Test
No build step — pure stdlib. From the repo root (mirrors .github/workflows/ci.yml):
# Syntax check (no deps)
python -m py_compile skills/rosetta/scripts/collect.py skills/rosetta/scripts/decisions.py \
skills/rosetta/scripts/ingest.py skills/rosetta/rosetta_cli.py
# Unit tests — agent resolvers/parsers, loop CLIs, ingest (run from skills/rosetta so `tests` imports)
( cd skills/rosetta && python -m unittest discover -s tests -v )
# Decision library must stay well-formed (anti-hallucination integrity gate)
python skills/rosetta/scripts/decisions.py --root skills/rosetta/decisions validate --integrity
Coverage runs as a separate CI job (kept out of the test job so that stays zero-dep): from
skills/rosetta/, coverage run -m unittest discover -s tests && coverage report — fail_under = 90
is enforced via .coveragerc. CI runs the suite on Python 3.9 and 3.12; keep both green.
Optional: pip install -e skills/rosetta puts rosetta on PATH; python3 skills/rosetta/tests/live_smoke.py
runs non-asserting counts against your real machine.
Architecture
skills/rosetta/rosetta_cli.py is a thin dispatcher that shells each subcommand to a deterministic
script under skills/rosetta/scripts/:
collect.py— resolve + normalize transcripts for one project: 18 per-agent resolvers + parsers, a discovery sweep, crash-safe atomic writes, and a processed-session ledger.--all-projectsemits a machine-wide discovery index.decisions.py— ADR/PDR/BDR records:new | index | validate | integrity | staleness | search | get | supersede | resolve | coverage. Owns numbering, the index, and validation.ingest.py— external decisions/signals →Status: Proposeddrafts (never auto-accepted).gates.py/preflight.py— local provenance/evidence gates;preflightalso embeds RA1's structural readiness report.runs.py— isolated loop-run ledger underloop-runs/.harness.py— allowlisted harness-doc export.drift.py— decision-freshness JSON.
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 · 84 lines · 1,261 tokens per session scan A 8a06f4093d38
rosetta AGENTS.md is an instructions file published in the GitHub repository tjboudreaux/rosetta (2 stars, last pushed 1mo ago), licensed MIT. It adds 1,261 tokens to every session, about $0.0063 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
EGC CLAUDE.md
Claude Code instructions for Fmarzochi/EGC, covering egc: session memory protocol, at the start of every session, at the end of every session, parallel sessions and where state is stored.
EGC GEMINI.md
Gemini CLI instructions for Fmarzochi/EGC, covering egc: session memory protocol, at the start of every session, at the end of every session, where state is stored and mcp servers required.
EGC copilot-instructions.md
Copilot instructions for Fmarzochi/EGC, covering copilot instructions for egc, repository conventions, readme and translations, code review priorities and egc project memory.
cairn AGENTS.md
Instructions for isaacriehm/cairn, covering cairn — agent orientation, what this project is, document index, hard rules and workspace layout.
adr-kit CLAUDE.md
Instructions for rvdbreemen/adr-kit, covering adr kit, backlog.md, werken in deze repo and graphify.
adr-kit AGENTS.md
Instructions for rvdbreemen/adr-kit, covering adr-kit agent instructions, task management, repo conventions, verification and adr kit.