Borrowing it
Nothing to install: this file belongs to adhi-jp/minecraft-modding-mcp. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/adhi-jp/minecraft-modding-mcp/main/AGENTS.mdgit clone --depth 1 https://github.com/adhi-jp/minecraft-modding-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/adhi-jp/minecraft-modding-mcp/agents-md)<a href="https://agentmods.dev/instructions/adhi-jp/minecraft-modding-mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/adhi-jp/minecraft-modding-mcp/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/instructions/adhi-jp/minecraft-modding-mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/adhi-jp/minecraft-modding-mcp/agents-md.svg" alt="Reviewed on agentmods" width="80" 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 | $0.03778 | $0.03778 |
| Opus 5.5 | $0.01511 | $0.01511 |
| Sonnet 5.5 | $0.00756 | $0.00756 |
| Haiku 4.5 | $0.00378 | $0.00378 |
Grade A, and why
minecraft-modding-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 — 119 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Purpose and Scope
- This file defines mandatory operating rules for agents working in this repository.
- When tradeoffs conflict, prioritize public API compatibility, verification evidence, and release safety.
Core Release Policy
- From version 1.0.0 onward, the public MCP tool surface follows semantic versioning. Breaking changes require a major version bump.
- Do not bump any version unless explicitly instructed by the user. Version bump timing is decided by the user.
Public MCP Contract (MUST)
- Treat the public contract as compatibility-bound: tool names, input parameter names/types/requiredness, and response envelope shape.
- Public MCP tool responses MUST use the standard envelope:
{ result?, error?, meta }. - Input validation failures MUST map to
ERR_INVALID_INPUTand MUST NOT be reported asERR_INTERNAL. - Do not add compatibility aliases for renamed public tools or parameters unless explicitly requested by the user and documented with a removal plan.
Change Coupling Rules (MUST)
- Any user-visible or public API change MUST update
CHANGELOG.mdand the relevant tests in the same change set. README.md,docs/README-ja.md, anddocs/tool-reference.mdMUST be updated in the same change set when the change touches something they already document — a tool, parameter, response field, error code, environment variable, or setup step. The Japanese overview needs an edit only where it already documents the item. All three are shipped in the package, so a stale statement there reaches users. A change with no counterpart in a file does not need an edit to it.- Do not defer documentation or test updates to follow-up commits.
Verification Gate (MUST)
- Before claiming completion for production changes, run the verification appropriate to what actually changed — not reflexively the full matrix below, and never less than what the change touches.
- Run the checks of every row the change touches. When unsure whether something reads a changed file, run the broader check.
src/**,tsconfig.json,pnpm-lock.yaml, or thepackage.jsonfieldstype,dependencies,devDependencies, orpackageManager:pnpm checkandpnpm test.tests/**,scripts/**, or thepackage.jsonscriptsfield:pnpm test, the whole suite: contract tests scan all ofsrc/,tests/, andscripts/, so an edit can fail a test other than the edited one.pnpm checkdoes not cover these; it type-checks onlysrc/**/*.ts. Also run each changedscriptsentry once, with representative arguments where it needs them, and stop a long-running one such asdevonce it has started.README.md,docs/**,CHANGELOG.md,.github/workflows/**, and any file orpackage.jsonfield no other row names: only the test files that read it, run withpnpm test:file <path...>(pnpm testtakes no file argument). Find them withgrep -rlF '<base name>' tests --include='*.test.ts'and run every match; some tests build the path, so search the base name, not the full path. Also runpnpm check:changelogwhenCHANGELOG.mdor thepackage.jsonversionchanged.AGENTS.md, or onlypackage.jsonfields that no test,src/file, or script reads (description,keywords,author,license,repository,bugs,homepage): nopnpm checkor test run; confirm that a changedpackage.jsonstill parses withnode -e "require('./package.json')".
- An edit that changes only comments runs no
pnpm checkor test, whichever row its path falls under. This applies even after a run on the earlier bytes; a hook-requested comment rewording is no exception. Directive comments (@ts-expect-error,@ts-ignore,/// <reference>, lint or coverage pragmas) count as code, not comments. One exception: sometests/contracts/scanners also match comment text, such astests/contracts/no-sdk-private-request-handler-access.test.ts. When a new comment contains a literal one of them forbids, run that scanner withpnpm test:file. - A release commit is not covered by the narrower rows: Release Safety requires
pnpm teston the Node major the publish workflow pins before the release commit is created. - When MCP transport/tool registration or manual workflows change, also run
pnpm test:manual:stdio-smokewhen environment permits. - For search, index, or performance-sensitive changes, also run
pnpm test:perf. - Keep a change when its concrete user or developer value outweighs implementation, verification, future maintenance, and compatibility costs. Pure cleanup that removes dead or duplicated code needs behavioral-equivalence evidence, not a minimum speed improvement. Do not add speculative complexity for an unproven benefit; when the balance is unclear, report a candidate rather than implement it.
- Support speed claims with representative before/after measurements proportionate to the claim and risk; record the workload, configuration, and results, including absolute as well as relative differences where useful. There is no fixed percentage threshold. Judge memory, timeout, and resource-correctness fixes against the actual failure or limit they address, not a latency percentage. Existing checks above remain required for the affected scope.
- Do not re-run a verification command solely to re-confirm a result already produced for the current, unchanged tree. If a later run — your own or a delegated worker's — reports an outcome that contradicts a fresh prior run with no tree change in between, diagnose the discrepancy before re-running the full suite again; a concurrent invocation sharing the same working tree is a common non-code cause.
- Do not claim "done", "fixed", or "passing" without fresh command output evidence.
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 Changed · +3 lines · +355 tokens per session 4199cf20e7b1
- 22d ago Changed · +5 lines · +177 tokens per session 8ef3f8948089
- 1mo ago Changed · +1 lines · +214 tokens per session b9ccbfc2dd34
- 1mo ago Changed · +2 lines · +233 tokens per session b14652b888be
- 2mo ago Changed · +5 lines · +247 tokens per session 942d9d7bd84f
- 2mo ago First seen · 103 lines · 2,552 tokens per session scan A f5826f2e541c
minecraft-modding-mcp AGENTS.md is an instructions file published in the GitHub repository adhi-jp/minecraft-modding-mcp (8 stars, last pushed 4d ago), licensed MIT. It adds 3,778 tokens to every session, about $0.0151 per session on Opus 5.5. 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.
minecraft-modding-mcp AGENTS.md compared with instructions from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| minecraft-modding-mcp AGENTS.mdadhi-jp/minecraft-modding-mcp | Instructions file | 8 | 4d ago | MIT | A | 3,778 | Codex, OpenCode |
| vscode buildNext.instructions.mdmicrosoft/vscode | Instructions file | 195k | today | MIT | A | 9,719 | GitHub Copilot |
| next.js AGENTS.mdvercel/next.js | Instructions file | 143k | yesterday | MIT | A | 7,461 | Codex, OpenCode |
| vscode oss-third-party-notices.instructions.mdmicrosoft/vscode | Instructions file | 195k | today | MIT | A | 5,649 | GitHub Copilot |
| langchain AGENTS.mdlangchain-ai/langchain | Instructions file | 149k | today | MIT | A | 4,680 | Codex, OpenCode |
| vscode copilot-instructions.mdmicrosoft/vscode | Instructions file | 195k | today | MIT | A | 3,093 | GitHub Copilot |
| vscode ai-customization.instructions.mdmicrosoft/vscode | Instructions file | 195k | today | MIT | A | 1,907 | GitHub Copilot |