agentmods.dev

Scout: Instructions file for Claude Code

CLAUDE.md

Scout CLAUDE.md is an instructions file for Claude Code from Raven-Scout/Scout. It costs 1,728 tokens per session, scanned A, original, MIT.

Development instructions for Scout, a macOS desktop app that works with a real person's information vault. They define how test data must be anonymized before entering the public repository.

In plain words
What is it for?
Use them when adding fixtures, inline test data, or documentation to Scout and its related iOS and plugin repositories.
Why use it?
They prevent private names, company details, links, and identifiers from being copied into public fixtures or tests.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md. Also seen: mentions .claude-plugin; mentions CLAUDE.md; mentions Claude Code.

This is Raven-Scout/Scout's own configuration. It tells Claude Code how to work on Scout itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything Scout configures →

Runs only inside a plugin — its command needs a path that Claude Code sets for a plugin’s own hooks and for nothing else, and the catalogue could not identify which plugin ships it.

not rated 20repo changed 7d ago A scan 1,728 tokens original MIT
View source ↗ Raven-Scout/Scout raven-scout.github.io
Reuse

Borrowing it

Nothing to install: this file belongs to Raven-Scout/Scout. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/Raven-Scout/Scout/main/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/Raven-Scout/Scout

Made for: Claude Code.

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 Scout CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/raven-scout/scout/claude-md/github.svg)](https://agentmods.dev/instructions/raven-scout/scout/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/raven-scout/scout/claude-md"><img src="https://agentmods.dev/badge/instructions/raven-scout/scout/claude-md/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for Scout CLAUDE.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/raven-scout/scout/claude-md"><img src="https://agentmods.dev/badge/instructions/raven-scout/scout/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 1,728 This file is loaded in full into every session.
When invoked 1,728 The same file — it is already loaded in full.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.01728 $0.01728
Opus 5.5 $0.00691 $0.00691
Sonnet 5.5 $0.00346 $0.00346
Haiku 4.5 $0.00173 $0.00173

Measured 7d ago against content hash 764af33f3740, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

Scout 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 7d 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.

CLAUDE.md · 97 lines

How it starts

The opening of the file, as written. The whole thing — 97 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

Working notes for AI agents in the Scout monorepo.

This repo holds two shipping artifacts

Path Artifact Read next
plugin/ The Scout Claude Code plugin + Python engine (scoutctl) plugin/CLAUDE.md
apps/macos/ The Scout macOS menu-bar app (SwiftUI) apps/macos/CLAUDE.md

Work in the subtree that owns the change, and read that subtree's CLAUDE.md first. The two have different languages, test runners, and release flows.

Rules that span both subtrees

  • .claude-plugin/marketplace.json lives at the REPO ROOT, not in plugin/. claude plugin marketplace add reads it from there. plugin.json — the canonical version — lives at plugin/.claude-plugin/plugin.json. Its scout entry pins a git-subdir source to the stable branch (#362), so installs get the latest release, never main's tip. Only release.sh finalize moves stable. Never point the entry back at "./plugin", and never push stable by hand except to run the recovery command finalize prints.
  • plugin/.claude-plugin/marketplace.json makes plugin/ its own marketplace root: the root manifest with only the scout entry, sourced "./". Hosts that serve a checkout and can't follow git-subdir (Agnes) read it from a vX.Y.Z tag, never main (see plugin/README.md), and Scout.app's engine bundle ships it (bundle-engine.sh fails if it differs from what it would generate). Its version is a release version file like the others; when the root entry changes, change this one to match, keeping "./".
  • Everything the plugin needs at runtime must stay under plugin/. CLAUDE_PLUGIN_ROOT resolves to the materialized plugin/ subtree only; a plugin file that reaches outside it is broken for every installed user, and nothing in this repo will tell you.
  • One version, one tag, one release. Scout's app and plugin share one version X.Y.Z. It lives in plugin/.claude-plugin/plugin.json, the root marketplace.json, plugin/.claude-plugin/marketplace.json, plugin/engine/pyproject.toml, plugin/engine/scout/__init__.py, and every MARKETING_VERSION in apps/macos/Scout.xcodeproj; versioning check (CI: contract) fails on drift. From v0.16.0 every release is ONE tag vX.Y.Z, cut with scripts/release.sh prepare then finalize; release candidates are vX.Y.Z-rc.N, cut with rc from release/vX.Y.Z. Never create app/v* or plugin/v* tags (sole exception before v0.16.0: an urgent plugin/v0.15.x patch via plugin/scripts/release-plugin.sh); the existing ones (app/v0.1.0–app/v0.14.0, plugin/v0.14.0, plugin/v0.15.0) are history. The app was never released as 0.15.0. Bare v0.4.0–v0.13.0 are the plugin's pre-monorepo releases. Agents never run finalize/rc for real.
  • Issue and PR numbers come from two repos. This repo is the former Raven-Scout/scout-plugin, renamed Raven-Scout/Scout when the app moved in, so its own #N numbering is scout-plugin's. The old app repo is Raven-Scout/scout-app-legacy (archived). A bare #N in apps/macos/ files or in a pre-move app commit subject, and any Raven-Scout/Scout#N written before the move, means Raven-Scout/scout-app-legacy#N. Link old app items there; never rewrite them to a bare #N here.
  • Contract artifacts are generated, never hand-edited. The connector roster and schedule snapshot each have ONE canonical file under plugin/engine/scout/, generated from its .yaml, and every client copy under apps/*/ is a byte copy of it. To change one, from plugin/engine: edit scout/connectors.yaml (or scout/defaults/schedule.yaml), regenerate the canonical file with .venv/bin/python -m scout.scripts.connectors_snapshot --no-also-write-app-fixture (or schedule_snapshot), then from the repo root cp it over every client copy contract.yml checks. Always pass --no-also-write-app-fixture — the default also writes into a sibling ~/scout-app checkout outside this repo, and never into apps/. contract.yml fails the build if a copy drifts. See docs/superpowers/specs/2026-09-03-monorepo-consolidation-design.md §7.
  • The parser corpus has one canonical copy too: plugin/engine/tests/fixtures/contract/parser-corpus.json. Edit it, set EXPECTED_SHA256 (plugin/engine/tests/unit/test_parser_corpus_checksum.py) and canonicalSHA256 (apps/macos/ScoutTests/ActionItems/ParserContractTests.swift) to its new shasum -a 256, then cp it to apps/macos/ScoutTests/Fixtures/parser-corpus.json — all in one commit. (scout-iOS-app vendors a copy in its own repo; see apps/macos/CLAUDE.md.)
  • The doctor run-health fixture has one canonical copy too: plugin/engine/tests/fixtures/contract/doctor-run-health.json, generated — never hand-edited — from the vault builder in plugin/engine/tests/unit/test_doctor_run_health_contract.py: from plugin/engine run env SCOUT_REGEN_CONTRACT=1 .venv/bin/python -m pytest tests/unit/test_doctor_run_health_contract.py, then from the repo root cp it to apps/macos/ScoutTests/Fixtures/doctor-run-health.json and update the pins in apps/macos/ScoutTests/Engine/DoctorReportTests.swift — all in one commit. contract.yml byte-compares the two copies.
  • Never edit a *.snapshot.json by hand. That is what broke the shipped connector roster for four months.
  • Plugin history was absorbed with git subtree, SHAs preserved. git blame plugin/<file> crosses the merge to the original commits, but git log -- plugin/<file> stops at the merge commit — for a plugin file's earlier history run git log <subtree-merge>^2 -- <path without the plugin/ prefix> (the first subtree merge is the commit titled "absorb scout-plugin into plugin/").

Read the full file on GitHub · 97 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. 7d ago Changed · +7 lines · +154 tokens per session 764af33f3740
  2. 16d ago Changed · +10 lines · +166 tokens per session 287691bf9a31
  3. 17d ago Changed · +9 lines · +300 tokens per session fb3383247b28
  4. 23d ago Changed · -12 lines · -233 tokens per session scan B → A 9fdbbe3a419a
  5. 29d ago Changed · +37 lines · +622 tokens per session scan A → B b77903c5fceb
  6. 1mo ago First seen · 46 lines · 719 tokens per session scan A dd4e05bc1977

Subscribe to this mod's changes

Scout CLAUDE.md is an instructions file published in the GitHub repository Raven-Scout/Scout (20 stars, last pushed today), licensed MIT. It adds 1,728 tokens to every session, about $0.0069 per session on Opus 5.5. 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.

Autonomous AgentsBuilding agentsClaude CodeWriting and docsagenticai-agentanthropicclaude-code-plugin
Related

Other instructions, from other repositories

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 9,715 tokens

next.js AGENTS.md

AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,461 tokens

vscode oss-third-party-notices.instructions.md

Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).

microsoft/vscode · 5,262 tokens

langchain AGENTS.md

AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.

langchain-ai/langchain · 4,680 tokens

vscode copilot-instructions.md

Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).

microsoft/vscode · 2,731 tokens

deepseek-harness AGENTS.md

AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.

deepseek-ai/deepseek-harness · 4,261 tokens