iago CLAUDE.md

iago CLAUDE.md is an instructions file for coding agents from drakulavich/iago. It costs 1,769 tokens per session, scanned A, original, MIT.

A repository guide for coding agents working on Iago. Iago adds a Mermaid diagram—a text-based flow, sequence, class, or database-relationship drawing—to a GitHub pull request review comment.

In plain words
What is it for?
Use it to understand Iago’s structure, install or invoke it, follow its Claude Code plugin setup, and run the project’s supported workflow.
Why use it?
It gives agents the project overview and critical rules they need before changing the code, including which runtime and tools the project expects.

Instructions file

Install

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.

agentmods
npx agentmods add instructions/drakulavich/iago/claude-md
Clone the repo
git clone --depth 1 https://github.com/drakulavich/iago

Wrote 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.

agentmods badge for iago CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/drakulavich/iago/claude-md.svg)](https://agentmods.dev/instructions/drakulavich/iago/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/drakulavich/iago/claude-md"><img src="https://agentmods.dev/badge/instructions/drakulavich/iago/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 1,769 This file is loaded in full into every session.
When invoked 1,769 The same file — it is already loaded in full.
Security scan A 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.01769 $0.01769
Opus 5 $0.00885 $0.00885
Sonnet 5 $0.00354 $0.00354
Haiku 4.5 $0.00177 $0.00177

Measured 4d ago against content hash c59eb94c3c26, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

iago CLAUDE.md scanned grade A with 1 finding 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

Full install options (`curl`, `--target`, `--version`, `doctor`) and end-user
CLAUDE.md · 134 lines

How it starts

The opening of the file, as written. The whole thing — 134 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

Single source of truth for every coding agent (Claude Code, Codex, Cursor, Aider, Copilot, Gemini, …) working on or with this repo. When this file and any other doc disagree, this file wins.

Project overview

Iago is a skill for AI coding agents that appends a Mermaid diagram (sequence / flow / class / entity-relation) to a GitHub pull request's /review comment. The host agent draws the diagram with the LLM it's already running; Iago's helper posts it via gh. The diagram type is auto-detected from the diff (overridable).

It is skill-first, distributed three ways:

  • skill folders (iago/, squawk/) copied into an agent's skills dir,
  • a Claude Code plugin (.claude-plugin/),
  • a TypeScript installer CLI (cli/, published as @drakulavich/iago).

Critical rules (read before changing anything)

  • No GitHub Action. Iago used to ship a server-side Action with its own LLM-provider SDKs + a heuristic fallback. That was removed deliberately. Do not reintroduce an Action, an API-key/provider path, or a heuristic diagram generator — the diagram is always drawn by the host agent's LLM.
  • Runtime is bun + gh, nothing else. The helper (iago/scripts/post.ts) is TypeScript run by bun; GitHub writes go through authenticated gh. The old install.sh (shell) and sanitize_mermaid.py (Python) were removed — don't bring shell/python back into the skill runtime.
  • The skill folder stays dependency-free. iago/scripts/*.ts may import only Node built-ins (node:*) + each other. Never add a package.json or node_modules under iago/ — the folder is copied verbatim into agent dirs.
  • Mermaid: never use reserved keywords as participant/node ids. loop, alt, opt, par, note, end, activate (case-insensitive) break GitHub's renderer. (Incident: a participant Loop made sequence.md fail to render.) Likewise never start an unquoted flowchart node label with @ — Mermaid lexes [@ as edge-ID/shape syntax (incident: N[@utils/utils -> …] killed a posted diagram; sanitize.ts now auto-quotes these). Repo example diagrams are parse-validated by cd cli && bun test (cli/tests/mermaid-validation.test.ts); field rendering incidents go into cli/tests/fixtures/mermaid/{valid,invalid}/ verbatim.
  • post.ts gh PATCH uses -f (raw-field), not -F. gh api -F/--field treats a leading @ as a file-read; comment bodies starting with @mention would break. -f/--raw-field is the safe static-string flag. Keep it.
  • Never commit local dev state. .omc/, .claude/, .codegraph/ are gitignored. Do not git add -A — stage files explicitly. (These dirs were once swept into a PR by git add -A.)
  • Versions stay aligned. cli/package.json, .claude-plugin/plugin.json, and .claude-plugin/marketplace.json (both version fields) move together.
  • No eval suite. Diagram quality is validated by dogfooding (run /iago on real PRs when the rubric changes), not a formal benchmark — a deterministic classifier no longer exists in skill-first, and an LLM-judge harness is overkill here. Don't rebuild evals/.

Read the full file on GitHub · 134 lines

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.

  1. 4d ago First seen · 134 lines · 1,769 tokens per session scan A c59eb94c3c26

Subscribe to this mod's changes

iago CLAUDE.md is an instructions file published in the GitHub repository drakulavich/iago (2 stars, last pushed 2mo ago), licensed MIT. It adds 1,769 tokens to every session, about $0.0088 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.