Borrowing it
Nothing to install: this file belongs to Abinesh-L/claude-crusts. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/Abinesh-L/claude-crusts/main/CLAUDE.mdgit clone --depth 1 https://github.com/Abinesh-L/claude-crustsWrote 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/abinesh-l/claude-crusts/claude-md)<a href="https://agentmods.dev/instructions/abinesh-l/claude-crusts/claude-md"><img src="https://agentmods.dev/badge/instructions/abinesh-l/claude-crusts/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.
<a href="https://agentmods.dev/instructions/abinesh-l/claude-crusts/claude-md"><img src="https://agentmods.dev/badge/instructions/abinesh-l/claude-crusts/claude-md.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.07152 | $0.07152 |
| Opus 5 | $0.03576 | $0.03576 |
| Sonnet 5 | $0.01430 | $0.01430 |
| Haiku 4.5 | $0.00715 | $0.00715 |
Grade B, and why
claude-crusts CLAUDE.md scanned grade B 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 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
- **hooks.ts**: Claude Code hook integration — reads/writes `~/.claude/settings.json` hooks, manages `~/.claude-crusts/config.json` toggle state How it starts
The opening of the file, as written. The whole thing — 235 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md — claude-crusts: Context Window Analyzer for Claude Code
Identity
You are working on claude-crusts (the npm package name and CLI binary). The framework is called CRUSTS (always uppercase). It analyzes Claude Code's context window by classifying every token into 6 categories: Conversation, Retrieved, User, System, Tools, State. It tells users WHY their context is filling up and WHAT to do about it. Fully offline, zero API calls.
CLI Commands
Binary name: claude-crusts. Entrypoint: src/index.ts.
Slash command: /crusts inside Claude Code (via .claude/commands/crusts.md) — runs analyze and gives actionable advice.
claude-crusts analyze [session-id] — 6-category breakdown + waste detection + recommendations
claude-crusts waste [session-id] — Detailed waste report with per-file analysis + top 5 consumers
claude-crusts fix [session-id] — 3 pasteable prompt blocks (session, CLAUDE.md, /compact)
claude-crusts timeline [session-id] — Message-by-message context growth with compaction markers
claude-crusts list — All discovered sessions (age, size, project)
claude-crusts compare <session-a> <session-b> — Side-by-side comparison with per-category deltas + insights
claude-crusts lost [session-id] — What was lost during compaction (files, conversations, tools, instructions)
claude-crusts watch [session-id] — Live-monitor a session with compact dashboard
claude-crusts report [session-id] — Generate standalone report (HTML or Markdown)
claude-crusts calibrate — Cross-reference against /context output
claude-crusts trend — Cross-session trends (sparkline, averages, direction)
claude-crusts tui [session-id] — Interactive REPL shell with tab completion + clipboard copy
claude-crusts status [session-id] — One-line context health (fast path, used by hooks)
claude-crusts hooks enable|disable|status — Manage Claude Code hook integration
claude-crusts statusline — Render statusline glyph (reads Claude Code JSON on stdin)
claude-crusts statusline install|uninstall|status — Manage Claude Code statusline integration
claude-crusts optimize [session-id] — Ranked actionable fixes with token-savings ROI (--apply to write changes)
claude-crusts models [session-id] — Per-session model usage snapshot (one row per contiguous same-model segment)
claude-crusts doctor — Sanity-check the install (Claude Code, hooks, statusline, backups)
claude-crusts diff [session-id] --from <n> --to <n> — Intra-session diff: per-category delta between two message cutpoints
claude-crusts trend --format csv [--output <path>] — Export trend history as CSV for external analysis
claude-crusts hooks auto-inject enable|disable|status|log — Manage hook-triggered fix injection (UserPromptSubmit); `log` shows history at ~/.claude-crusts/auto-inject.log
claude-crusts auto-inject — Internal hook target (reads stdin; emits additionalContext when hot)
claude-crusts bench compact [session-id] — Tail the session JSONL for a /compact event; write before/after delta to JSON
claude-crusts bench reextract <result.json> — Re-run the file-ref extractor against an existing bench result (no recompact)
claude-crusts bench compare <a.json> <b.json> — Diff two bench-compact results (survivor file-ref set + token metrics)
claude-crusts completion <bash|zsh|pwsh> — Emit a shell completion script (TAB completes subcommands + session IDs)
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 Changed · +7 lines · +1,713 tokens per session 03ecfd48a8dc
- 11d ago First seen · 228 lines · 5,439 tokens per session scan B b4ab5aceaf5f
claude-crusts CLAUDE.md is an instructions file published in the GitHub repository Abinesh-L/claude-crusts (83 stars, last pushed 10d ago), licensed MIT. It adds 7,152 tokens to every session, about $0.0358 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
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.
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
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).
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
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.