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/orange91pku/unity-ai-project-brain/agents-mdgit clone --depth 1 https://github.com/orange91pku/unity-ai-project-brainWrote 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/orange91pku/unity-ai-project-brain/agents-md)<a href="https://agentmods.dev/instructions/orange91pku/unity-ai-project-brain/agents-md"><img src="https://agentmods.dev/badge/instructions/orange91pku/unity-ai-project-brain/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 | $0.00568 | $0.00568 |
| Opus 5 | $0.00284 | $0.00284 |
| Sonnet 5 | $0.00114 | $0.00114 |
| Haiku 4.5 | $0.00057 | $0.00057 |
Grade A, and why
unity-ai-project-brain 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 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 — 67 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AI agent guide
Mission
Unity Project Brain is a read-only context engine for Unity repositories. It joins code intelligence, Unity GUID/fileID identity, serialized scene and prefab relationships, optional Roslyn symbols, runtime evidence, and Git history so AI agents can reason about a project before proposing changes.
Safe boundary
- Treat the configured Unity repository as private unless the operator says otherwise.
- Do not commit local project configuration, indexes, databases, caches, credentials, or absolute machine paths.
- Do not mutate Unity source through Project Brain. Mutation belongs in an external tool and its result must pass Unity verification.
- Preserve unresolved and ambiguous relationships as explicit evidence; never invent a link to make an answer look complete.
- Keep the Context API loopback-only and MCP read-only unless the operator deliberately changes that boundary.
Authoritative sources
- Git working tree and commits: code content and history.
- Unity
.metafiles: asset GUID identity. - Text-serialized
.unity,.prefab, and.assetfiles: object, component, hierarchy, and serialized-reference identity. - Roslyn-compatible symbol catalogs: optional code symbol identity.
- Attributed verification reports: acceptance evidence.
- GitNexus: rebuildable code graph and semantic retrieval index.
- Project Brain SQLite graph: rebuildable Unity identity and evidence index.
Entry points
src/cli.mjs: GitNexus orchestration and scoped MCP descriptors.project_brain/unity_graph.py: Unity graph build, validate, and query.project_brain/context_api.py: read-only CLI and loopback HTTP API.project_brain/change_envelope.py: versioned change contracts.schemas/: public interoperability contracts.docs/ARCHITECTURE.md: system boundaries and data flow.docs/GETTING_STARTED.md: reproducible setup.docs/UPM_INSTALL.md: no-clone Unity package distribution.Packages/com.heycici.project-brain/: read-only Unity Editor bridge.
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 · 67 lines · 568 tokens per session scan A d9e29eca11a9
unity-ai-project-brain AGENTS.md is an instructions file published in the GitHub repository orange91pku/unity-ai-project-brain (2 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 568 tokens to every session, about $0.0028 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
repobrain copilot-instructions.md
Copilot instructions for study8677/repobrain, covering github copilot bootstrap instructions and hard rule — query the repobrain hub first.
ai AGENTS.md
AGENTS.md instructions for vercel/ai, covering agents.md, project overview, repository structure, key directories and core package dependencies.
GameDesignOS AGENTS.md
Instructions for DY-2026/GameDesignOS, covering gamedesignos agent 入口, 默认接法, agent 执行规则, 什么时候用整个项目 and 每个 skill 单独怎么用.
moorestech AGENTS.md
AGENTS.md instructions for moorestech/moorestech, covering コーディングの指針, qa（必須）, 互換性とパフォーマンス, 既知の制約（設計上の割り切り） and regionの活用.
OwnPilot AGENTS.md
Instructions for ownpilot/OwnPilot, covering ownpilot, architecture, key patterns, commands and tech stack.
codescout CLAUDE.md
Claude Code instructions for mareurs/codescout, covering codescout, development commands, bug tracking, session intelligence trackers and querying active trackers (librarian).