Pinvou/pinvou-agent is an open-source desktop workspace where an AI agent helps with work, visual design, and software development. People use it to work with files and knowledge, create editable visual artifacts, connect coding agents to projects, and extend the workspace with tools, MCP servers, skills, and workflows; the catalogue skills add capabilities to that environment.
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/pinvou/pinvou-agent/agents-mdgit clone --depth 1 https://github.com/Pinvou/pinvou-agentWrote 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/pinvou/pinvou-agent/agents-md)<a href="https://agentmods.dev/instructions/pinvou/pinvou-agent/agents-md"><img src="https://agentmods.dev/badge/instructions/pinvou/pinvou-agent/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 5.1 | $0.01748 | $0.01748 |
| Opus 5 | $0.00874 | $0.00874 |
| Sonnet 5 | $0.00350 | $0.00350 |
| Haiku 4.5 | $0.00175 | $0.00175 |
Grade A, and why
pinvou-agent AGENTS.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 6d 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 — 88 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Pinvou Agent Project Conventions
Core constraints
0. Local private memory
- If
.codex-memory.mdexists at the repository root, read it before starting work. It is local private memory and must not be committed.
1. Development and pull request conventions
The contribution workflow — syncing origin/main, conflict-resolution expectations, commit-message format and DCO, PR language, and local/CI checks — is defined once in CONTRIBUTING.md. The rules below restate only what an agent working in this repository must apply to every task:
- Before the first development work for a new task and before creating its pull request, safely sync the latest
origin/mainand align every submodule with the parent repository gitlink. Do not repeatedly sync only becausemainadvances during continued development or review of the same task. Before merge, the Merge Queue runs applicable gates against the combined tree on the latestmain; manually rebase only when a real conflict or queue integration failure requires changing the branch. - Before the final commit and pull request creation, self-review requirement completeness, root cause and related cases, implementation boundaries, impact on existing functionality, exceptional states, and verification sufficiency. Passing tests does not replace this review.
- Fix and verify any in-scope gap found during self-review. Disclose and await a decision for out-of-scope changes, mutually exclusive approaches, unverified scenarios, or regression risks.
2. CodeWhale and fork boundaries
CodeWhale provides the foundation for model calls, streaming output, tool loops, Sessions, Skills, Commands, MCP, Hooks, and Compaction. Pinvou Agent does not reimplement these capabilities.
Place extensions according to these boundaries:
| Change type | Location |
|---|---|
| Domain Agent or tool composition | SKILL.md |
| External API or standalone capability | MCP server / connector |
| Model behavior guidance | Bundle instructions.md |
| UI, Tauri integration, or Engine configuration | pinvou3-app/ |
| Reusable foundation issue | CodeWhale, upstream first |
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.
- 6d ago First seen · 88 lines · 1,748 tokens per session scan A d5cdb17949a2
pinvou-agent AGENTS.md is an instructions file published in the GitHub repository Pinvou/pinvou-agent (1,625 stars, last pushed today), licensed MIT. It adds 1,748 tokens to every session, about $0.0087 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 2026-08-30.
Other instructions, from other repositories
Abu-Cowork AGENTS.md
AGENTS.md instructions for PM-Shawn/Abu-Cowork, covering abu (阿布) — ai desktop office assistant, project overview, tech stack, git workflow & development constraints and branches.
Abu-Cowork CLAUDE.md
Claude Code instructions for PM-Shawn/Abu-Cowork, a project described as: Open-source alternative to Claude Cowork — a local-first AI agent desktop app · multi-model · self-evolving skills · privacy-first · multi-Harness roadmap · DeepSeek Harness integration in progress.
Team-X CLAUDE.md
Instructions for Git-Rocky-Stack/Team-X, covering claude.md, inheritance from parent repositories, design system, key contacts and skill routing.
madcop AGENTS.md
AGENTS.md instructions for linmy666/madcop, covering agents.md — madcop agent 协作规约, ⚡ 0. 身份定义 (identity), ⚡ 1. 黄金法则: 优先调用 madcop/tools/ 中的现成工具, ⚡ 2. 消息优先级高于协议 (message priority over protocol) and ⚡ 3. 长效编程任务规范 (long-horizon programming).
rizzo-pii CLAUDE.md
Instructions for Rizzo-AI-Academy/rizzo-pii, covering claude.md, cos'è questo progetto, ambiente — vincoli critici e non ovvi, mappa della repo and tassonomia: 22 tag (dettaglio in tassonomiatag.md).
remnic AGENTS.md
AGENTS.md instructions for joshuaswarren/remnic, covering remnic - agent guide, architecture boundaries (non-negotiable), upstream references, adapter implementation rules and openclaw compatibility window.