campy CLAUDE.md

Repository instructions for Campy, a tool that displays animated ASCII pets in terminal coding agents. They explain its Claude Code integration, state-file communication, commands, and legacy ghost-pet plugin, which requires jq.

In plain words
What is it for?
Use them when developing or debugging Campy, its command-line tool, or its Claude Code and OpenCode integrations. They also cover setup and common commands.
Why use it?
They give an agent the project context needed to make consistent changes. They also document how the core code, agent adapters, hooks, and state storage fit together.

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/dropdevrahul/campy/claude-md
Clone the repo
git clone --depth 1 https://github.com/dropdevrahul/campy
Per session 3,286 This file is loaded in full into every session.
When invoked 3,286 The same file — it is already loaded in full.
Security scan B 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 $0.03286 $0.03286
Opus 5 $0.01643 $0.01643
Sonnet 5 $0.00657 $0.00657
Haiku 4.5 $0.00329 $0.00329

Measured 2d ago against content hash a4e218ca8738, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

campy 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 2d 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.

`[mcp_servers.campy]` in `~/.codex/config.toml`. **Cursor CLI** → `~/.cursor/mcp.json`.
CLAUDE.md · 214 lines

How it starts

The opening of the file, as written. The whole thing — 214 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Overview

campy ships animated ASCII terminal pets for CLI coding agents. The codebase is organized as a portable core/ + per-agent adapters architecture:

  1. core/ — single source of truth for all pet logic. No host-specific code. Modules: types.ts, frame-utils.ts, animation-engine.ts, gif-engine.ts, pets/{cat,hamster,ghost,robot}.ts + index.ts, theme.ts, personality.ts, happiness.ts, events.ts, store.ts, runtime.ts, render.ts. .opencode/plugins/ re-exports all modules from core/ (thin shims only).
  2. cli/campy.ts — the campy binary (runnable via bun). Subcommands: watch, attach, statusline, mcp, event, feed, play, pet, sleep, wake, switch, status, setup, install.
  3. adapters/pi/ — Pi coding-agent adapter (in-process TUI widget).
  4. adapters/claude-code/ — Claude Code adapter: short-lived hooks mutate the state file via the campy CLI; a campy watch pane or the statusline renders. Wires only two reactive hooks — PostToolUse (scoped to Edit|Write|MultiEdit|NotebookEdit|Bash) and Stop — plus a one-shot SessionStart; read-only tools never spawn a process. Contains hooks/, commands/, .claude-plugin/, and statusline.sh.
  5. adapters/mcp/ — MCP (stdio) server for agents whose native surface is MCP (Gemini CLI, Codex CLI, Cursor CLI). Watches the project dir → file_edited events, and exposes campy_* tools (status/feed/play/pet/switch) returning the pet as inline ASCII. dispatch() is unit-tested; runMcpServer() wires stdio.
  6. ghost-pet/ — legacy standalone Bash Claude Code plugin (ghost pet only); kept for backward compatibility until the adapters/claude-code/ adapter reaches full parity.

Installing into an agent

campy setup detects installed agents (via core/detect.ts) and wires each the native way; campy install <agent> does one. Native mechanism per agent: Claude Code → plugin marketplace (.claude-plugin/marketplace.json); the plugin auto-wires its hooks via adapters/claude-code/hooks/hooks.json and its statusline via the statusLine field in plugin.json (both ${CLAUDE_PLUGIN_ROOT}-relative). Or wire the same hooks + statusline manually in settings.json (printed by campy install claude-code). OpenCode/Pi → in-process extension. Gemini CLI~/.gemini/extensions/campy/ + gemini-extension.json. Codex CLI[mcp_servers.campy] in ~/.codex/config.toml. Cursor CLI~/.cursor/mcp.json. Aider.git/hooks/post-commit. For agents without a widget surface, campy attach spawns a campy watch pane (tmux/zellij/wezterm/kitty).

Read the full file on GitHub · 214 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. 2d ago First seen · 214 lines · 3,286 tokens per session scan B a4e218ca8738

Subscribe to this mod's changes

campy CLAUDE.md is an instructions file published in the GitHub repository dropdevrahul/campy (10 stars, last pushed 2mo ago), licensed MIT. It adds 3,286 tokens to every session, about $0.0164 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 2026-08-31.