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/xiaolai/claudepot-app/agents-mdgit clone --depth 1 https://github.com/xiaolai/claudepot-appWrote 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/xiaolai/claudepot-app/agents-md)<a href="https://agentmods.dev/instructions/xiaolai/claudepot-app/agents-md"><img src="https://agentmods.dev/badge/instructions/xiaolai/claudepot-app/agents-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 | $0.37982 | $0.37982 |
| Opus 5 | $0.18991 | $0.18991 |
| Sonnet 5 | $0.07596 | $0.07596 |
| Haiku 4.5 | $0.03798 | $0.03798 |
Grade D, and why
claudepot-app AGENTS.md scanned grade D with 2 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 today.
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.
`~/.claude/settings.json` and exists exactly while a grant is live. Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
`projects/<slug>/<uuid>.jsonl` also `rm -rf`s the session folder How it starts
The opening of the file, as written. The whole thing — 2,677 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Claudepot
Control center for Claude Code and Claude Desktop. Tauri 2 + Rust + React.
The seed was multi-account credential switching. The shipped surface is broader: accounts, projects, sessions, API keys, third-party integrations, agents, memory (CLAUDE.md files), usage/cost tracking, updates, service status, and notifications — all under one Tauri shell with tray + menubar integration.
The domain model in claudepot-core is five nouns: account, cli,
desktop, project, plus agent (scheduled headless claude -p
runs — the one noun added since the seed; see
claudepot-core::agent). Other surfaces are presentation layers
over those nouns and over CC's filesystem, not new domain types. See
.claude/rules/architecture.md for the noun-vs-surface distinction.
Scope discipline applies to the domain model (don't add nouns
casually); it does not cap what the UI can usefully expose.
Shared memory (dogfooding)
Claudepot indexes this repo's own Claude + Codex transcripts and
exposes them over MCP. The snippet below tells you which tools exist
and when to call them. It is generated — refresh with
claudepot mcp install-snippet --out .claude/claudepot-mcp-instructions.md;
never hand-edit it, and never duplicate it inline.
@.claude/claudepot-mcp-instructions.md
Build
cargo check --workspace # Rust
cargo build -p claudepot-cli # CLI binary
pnpm build # Frontend bundle
pnpm tauri dev # GUI in dev mode (hot reload)
pnpm tauri build --no-bundle # GUI release binary (no .dmg)
scripts/build-panel.sh # Remote panel → committed embed dir
scripts/build-panel.sh is separate because panel/ has its own
install and its output is committed — see "## Remote control". A
source change under panel/ that nobody rebuilt ships the previous
bundle with no error anywhere.
Test
cargo test --workspace # Rust
cargo xtask verify-cc-parity # CC settings-merge parity goldens (see parity-harness/README.md)
pnpm test # React (Vitest + RTL, jsdom)
pnpm test:coverage # React with coverage report
cd panel && pnpm check:render # the built remote panel actually mounts
pnpm check:classes # every className has a CSS rule behind it
pnpm check:a11y # every role="switch" has an accessible name
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.
- today Changed · +289 lines · +4,386 tokens per session 9c94eac6cc29
- 5d ago First seen · 2,388 lines · 33,596 tokens per session scan D 41f17222c1aa
claudepot-app AGENTS.md is an instructions file published in the GitHub repository xiaolai/claudepot-app (118 stars, last pushed today), licensed ISC. It adds 37,982 tokens to every session, about $0.1899 per session on Opus 5. A static security scan graded it D with 2 findings (reads agent configuration directories, recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
claude-code-settings copilot-instructions.md
Instructions for feiskyer/claude-code-settings, covering claude.md, environment setup, required dependencies, configuration and skills.
Get-Started-with-Web3 AGENTS.md
AGENTS.md instructions for beihaili/Get-Started-with-Web3, covering get started with web3 agent notes, 项目定位, ceo 运营目标, 常用命令 and ai-native 内容层.
vibe-coding-prompt-template tests.instructions.md
Instructions for KhazP/vibe-coding-prompt-template, a project described as: Templates and workflow for generating PRDs, Tech Designs, and MVP and more using LLMs for AI IDEs.
cherry-studio AGENTS.md
AGENTS.md instructions for CherryHQ/cherry-studio, a project described as: AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs.
the-pair AGENTS.md
Instructions for timwuhaotian/the-pair, covering ai agent guidelines for "the pair", project identity, tech stack, architecture rules and rust backend modules (src-tauri/src/).
ping-island AGENTS.md
AGENTS.md instructions for erha19/ping-island, covering agents.md, mission, start here, repo map and change routing.