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/alexeyperov/unity-open-mcp/agents-mdgit clone --depth 1 https://github.com/AlexeyPerov/Unity-Open-MCPWrote 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/alexeyperov/unity-open-mcp/agents-md)<a href="https://agentmods.dev/instructions/alexeyperov/unity-open-mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/alexeyperov/unity-open-mcp/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 | $0.01415 | $0.01415 |
| Opus 5 | $0.00707 | $0.00707 |
| Sonnet 5 | $0.00283 | $0.00283 |
| Haiku 4.5 | $0.00142 | $0.00142 |
Grade A, and why
Unity-Open-MCP 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 5d 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 — 53 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent rules
Rule scope and precedence
- Ancestor
AGENTS.mdrules always apply. A deeper file may add requirements or narrow an allowed choice for its subtree. It may contradict an ancestor only when that ancestor explicitly grants an exception. - For a cross-package change, satisfy the union of the rules governing every touched file; sibling rule files are not inherited automatically.
- Example: root permits tracked-doc updates and a package may narrow that to one owning API page. A package may not allow persisted-data migrations because root forbids them unless explicitly requested.
- Current local layers:
- Root (
AGENTS.md) — repository-wide safety, documentation, and maintainer process. packages/bridge/AGENTS.md— bridge transport, shipped embedded domains, registration, gates, and Editor UI.packages/verify/AGENTS.md— verify rules, fixes, issue codes, and capability catalog sync.packages/extensions/AGENTS.md— community/third-party extension packs and their template.mcp-server/AGENTS.md— MCP tools, routing, capabilities, and offline reads.hub/AGENTS.md— SvelteKit/Tauri UI, state, data, and platform neutrality.scripts/AGENTS.md— version sync, token-estimate codegen, and MCP test suites.skills/AGENTS.md— agent playbooks and client install-path manifest.demo/AGENTS.md— Unity integration fixture, manifest/testablessync.validation-suite/AGENTS.md— guided manual-validation app and scenarios.
- Root (
Universal contributor rules
- Specs (
specs/). This optional maintainer working tree is gitignored. Nevergit add, commit, or push anything underspecs/. Its absence in a public clone is not a blocker. - Migrations. Do not implement data migrations, compatibility shims, or upgrade paths for persisted data unless explicitly requested. Prefer simplifying storage and codecs over backward compatibility.
- No internal references in public surfaces. README files,
docs/, UI strings, and shipped skills must not exposespecs/paths, milestone/spec identifiers, execution-plan task numbers, or internal question citations.AGENTS.mdfiles and source comments may use an internal identifier only when it materially helps maintainers; prefer stable contract names even there. - No reference-project attribution outside
specs/. Names, repository paths, handles, and attribution to projects under/referencesmust not appear in source comments, checked-in docs, README files, skills, or UI strings. State contracts on their own merits. The only tracked exceptions are the deliberate comparison pagedocs/mcp-tools-comparison.mdand provenance logpackages/verify/EXTRACTION.md. - Docs are part of done. If a change affects public behavior, API contracts, architecture boundaries, or developer workflows, update tracked docs in
README.mdand/ordocs/in the same task. - Agent skill sync.
skills/unity-open-mcp/SKILL.mdis agent-facing guidance. The MCP package owns its synchronization with tool/capability/routing changes and the owning page indexed bydocs/api/mcp-tools.md; follow MCP server rules — Agent skill sync. Skill install paths come fromskills/client-paths.json, not per-package constants. - Docs layout and ownership.
- Root
README.mdstays short: intro, current feature set, quick links, and a Documentation section that is the docs index (no separatedocs/README.md). README.md's Documentation section is the docs index and must link every top-level doc, without duplicating links already provided by other README sections (Quick setup, Key features, Unity Hub Pro, Contributing). When a new top-level doc is added, link it from this section in the same task.docs/architecture.mdcovers repo structure and cross-package boundaries.docs/api.mdis the index for externally relevant API interfaces and contracts.
- Root
- Doc update scope rule. Edit only docs that match the changed area; avoid unrelated rewrites. If a new docs domain is introduced, add it to
README.md's Documentation section in the same task. - When docs updates can be skipped. Typos, formatting-only edits, comments-only changes, and internal refactors that do not alter behavior or contracts.
- Agent reporting requirement. If code changes but tracked docs are not updated, explicitly state why docs were not needed in the final handoff message.
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.
- 5d ago First seen · 53 lines · 1,415 tokens per session scan A 985e9073a40b
Unity-Open-MCP AGENTS.md is an instructions file published in the GitHub repository AlexeyPerov/Unity-Open-MCP (12 stars, last pushed 7d ago), licensed MIT. It adds 1,415 tokens to every session, about $0.0071 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
codeg AGENTS.md
AGENTS.md instructions for xintaofei/codeg, covering agents.md, 项目概述, 技术栈, 代码检查与测试(任务完成后进行必要的检查) and 前端.
agent-harness-kit AGENTS.md
AGENTS.md instructions for enmanuelmag/agent-harness-kit, covering agents.md — @cardor/agent-harness-kit, project, health check (run before making codebase changes), harness data (source of truth) and mcp tools (preferred).
open-vibe-island AGENTS.md
AGENTS.md instructions for Octane0411/open-vibe-island, covering agents, goal, required workflow, commit policy and safety rules.
open-vibe-island CLAUDE.md
Claude Code instructions for Octane0411/open-vibe-island, covering claude.md, project, architecture, build & run and dev app (open island dev.app).
codehydra CLAUDE.md
Instructions for stefanhoelzl/codehydra, covering claude.md, critical rules, no ignore comments, api/ipc interface changes and new boundary interfaces.
liteagents CLAUDE.md
Claude Code instructions for hamr0/liteagents, covering liteagents and documentation.