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/sprustonlab/claudechic/claude-mdgit clone --depth 1 https://github.com/sprustonlab/claudechicWrote 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/sprustonlab/claudechic/claude-md)<a href="https://agentmods.dev/instructions/sprustonlab/claudechic/claude-md"><img src="https://agentmods.dev/badge/instructions/sprustonlab/claudechic/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 | $0.02803 | $0.02803 |
| Opus 5 | $0.01401 | $0.01401 |
| Sonnet 5 | $0.00561 | $0.00561 |
| Haiku 4.5 | $0.00280 | $0.00280 |
Grade A, and why
claudechic 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 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 — 217 lines — stays where its author put it; the contents beside it link to each section on GitHub.
claudechic
A stylish terminal UI for Claude Code, built with Textual and wrapping the claude-agent-sdk.
Run
uv run claudechic
uv run claudechic --resume # Resume most recent session
uv run claudechic -s <uuid> # Resume specific session
Dev setup: uv sync --dev && source .venv/bin/activate (Linux/macOS).
File Map
claudechic/
├── __init__.py # Package entry, exports ChatApp
├── __main__.py # CLI entry point
├── agent.py # Agent class - SDK connection, history, permissions, state
├── analytics.py # PostHog analytics - fire-and-forget event tracking
├── agent_manager.py # AgentManager - coordinates multiple concurrent agents
├── app.py # ChatApp - main application, event handlers
├── awareness_install.py # claudechic-awareness install routine (~/.claude/rules/claudechic_*.md)
├── commands.py # Slash command routing (/agent, /shell, /clear, etc.)
├── compact.py # Session compaction - shrink old tool uses to save context
├── config.py # CONFIG (user-tier) + ProjectConfig (project-tier) loading
├── errors.py # Logging infrastructure, error handling
├── file_index.py # Fuzzy file search using git ls-files
├── formatting.py # Tool formatting, diff rendering (pure functions)
├── history.py # Global history loading from ~/.claude/history.jsonl
├── mcp.py # In-process MCP server for agent control tools
├── messages.py # Custom Textual Message types for SDK events
├── remote.py # HTTP server for remote control (live testing)
├── permissions.py # PermissionRequest dataclass for tool approval
├── profiling.py # Lightweight profiling utilities (@profile decorator)
├── sampling.py # CPU-conditional sampling profiler for high-CPU investigation
├── protocols.py # Observer protocols (AgentObserver, AgentManagerObserver)
├── sessions.py # Session file loading and listing (pure functions)
├── styles.tcss # Textual CSS - visual styling
├── theme.py # Textual theme definition
├── usage.py # OAuth usage API fetching (rate limits)
├── features/
│ ├── __init__.py # Feature module exports
│ ├── browser/ # Gated browser control MCP tools (experimental.browser)
│ │ ├── __init__.py # Public API (get_browser_tools, shutdown_browser, browser_enabled)
│ │ ├── session.py # BrowserSession - lazy Playwright-Chromium lifecycle
│ │ └── tools.py # browser_navigate/screenshot/click/type/scroll tools
│ └── worktree/
│ ├── __init__.py # Public API (list_worktrees, handle_worktree_command)
│ ├── commands.py # /worktree command handlers
│ └── git.py # Git worktree operations
├── workflows/ # Workflow orchestration ENGINE (Python code)
│ ├── __init__.py # Parser registration, public API
│ ├── engine.py # WorkflowEngine - phase state, advance checks, artifact_dir
│ ├── loader.py # ManifestLoader - 3-tier walk (package/user/project) + override resolution
│ ├── parsers.py # Section parsers for rules, hints, phases, etc.
│ ├── phases.py # Phase type, PhasesParser
│ └── agent_folders.py # Agent prompt assembly from workflow role dirs
├── checks/ # Check protocol (LEAF: stdlib only)
├── guardrails/ # Guardrail rules (LEAF: no upward imports)
├── hints/ # Hints pipeline (LEAF: stdlib only, no upward imports)
│ ├── __init__.py # Package marker
│ ├── engine.py # 6-stage hint evaluation pipeline
│ ├── parsers.py # Manifest section parser for hints YAML
│ ├── state.py # HintStateStore - persistence to .claudechic/hints_state.json
│ └── types.py # HintSpec, HintDecl, HintRecord, HintLifecycle, TriggerCondition
├── defaults/ # Bundled package-tier content (data, not Python)
│ ├── global/ # Always-active manifests
│ │ ├── rules.yaml # Guardrail rules
│ │ └── hints.yaml # Hint definitions
│ ├── workflows/ # Bundled workflow YAML directories (9 workflows)
│ │ ├── audit/
│ │ ├── cluster_setup/
│ │ ├── codebase_setup/
│ │ ├── git_setup/
│ │ ├── onboarding/
│ │ ├── project_team/
│ │ ├── tutorial/
│ │ ├── tutorial_extending/
│ │ └── tutorial_toy_project/
│ └── mcp_tools/ # MCP tool scripts (cluster dispatch, etc.)
├── context/ # Claude Code context docs (data, not Python)
│ ├── CLAUDE.md # User-facing quick reference (installed as ~/.claude/rules/claudechic_CLAUDE.md)
│ └── *.md # System docs (checks, guardrails, hints, workflows, etc.)
├── audit/ # Audit pipeline scripts
├── processes.py # BackgroundProcess dataclass, child process detection
├── screens/ # Full-page screens (navigation)
│ ├── chat.py # ChatScreen - main chat UI (default screen)
│ ├── diff.py # DiffScreen - review uncommitted changes
│ ├── session.py # SessionScreen - session browser for /resume
│ ├── chicsession.py # ChicsessionScreen - chicsession picker and workflow activation
│ ├── rewind.py # RewindScreen - checkpoint selection
│ ├── settings.py # SettingsScreen - in-app config editor (per /settings)
│ ├── disabled_workflows.py # DisabledWorkflowsScreen - per-(level, id) toggle subscreen
│ ├── disabled_ids.py # DisabledIdsScreen - per-(level, id) toggle subscreen for hints + rules
│ ├── welcome.py # WelcomeScreen - first-install onboarding
│ └── workflow_picker.py # WorkflowPickerScreen - workflow selection UI (with level badges)
└── widgets/
├── __init__.py # Re-exports all widgets for backward compat
├── prompts.py # All prompt widgets (Selection, Question, Model, Worktree)
├── base/ # Protocols and base classes
├── primitives/ # Low-level building blocks
├── content/ # Content display widgets
├── input/ # User input widgets
├── layout/ # Structural/container widgets (chat_view, sidebar, footer, indicators)
├── reports/ # In-page report widgets
└── modals/ # Modal screen overlays
tests/ # Test suite (pytest -- see Testing section)
docs/
└── dev/ # Developer documentation (from .ai-docs)
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 · 217 lines · 2,803 tokens per session scan A c46ac5f3f577
claudechic CLAUDE.md is an instructions file published in the GitHub repository sprustonlab/claudechic (2 stars, last pushed 8d ago), licensed MIT. It adds 2,803 tokens to every session, about $0.0140 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
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.
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.
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.
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).
next.js AGENTS.md
Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.