Borrowing it
Nothing to install: this file belongs to yongjip/mergetrain. 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/yongjip/mergetrain/main/AGENTS.mdgit clone --depth 1 https://github.com/yongjip/mergetrainWrote 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/yongjip/mergetrain/agents-md)<a href="https://agentmods.dev/instructions/yongjip/mergetrain/agents-md"><img src="https://agentmods.dev/badge/instructions/yongjip/mergetrain/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/yongjip/mergetrain/agents-md"><img src="https://agentmods.dev/badge/instructions/yongjip/mergetrain/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 | $0.01251 | $0.01251 |
| Opus 5.5 | $0.00500 | $0.00500 |
| Sonnet 5.5 | $0.00250 | $0.00250 |
| Haiku 4.5 | $0.00125 | $0.00125 |
Grade A, and why
mergetrain 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 9d 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 — 104 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
This repository contains mergetrain, a local deploy train for coding-agent
worktrees.
Agent operating rules
- Work on a task-specific branch and worktree.
- Commit all changes before enqueueing.
- Do not push deploy refs directly. For ordinary handoff, enqueue every named
finished branch in the requested order with
mergetrain enqueue --task <task> --branch <branch>and let mergetrain capture the exact SHAs. Stop after the last enqueue; asking to queue for validation does not authorize running validation. - Read
mergetrain status --jsonfirst. Task agents submit finished branches and stop; an authorized runner followsnext_actionfor validation, deployment, and recovery. A runner's wait action does not block submission. Usestatus --diagnoseonly for configuration, Git, runtime, or lock detail, andinspect <job-id>only for job evidence. - Use
--autoonly after explicit unattended-deploy approval. A bounded instruction to QA, deploy, verify, and finish end-to-end grants that approval for the named task and destination; do not ask for each opaque train ID. Auto jobs are bound to that Git destination and the approved gate, reuse, and verify policy; they must block if either identity changes. - Let one runner or daemon own merge and test. The default daemon may push and
verify only
--autojobs;daemon --validate-onlyhandles manual jobs but stops at a validated train and never deploys. - Fix blocked or failed work in the owning branch, commit a clean result, then
run
mergetrain retry <job-id>to replace that outcome with a freshly pinned job. Do not enqueue the same active branch again. Retry is for blocked/failed jobs only and does not itself authorize validation or deployment. - Do not delete or rewrite remote
refs/mergetrain/deploys/*; they are permanent recovery evidence. - Treat public product surface as an owner-evidence budget. Before
adding a CLI command or flag, config field, UI or network listener,
daemon/Hub behavior, MCP tool, recovery action, notification path, or reuse
control,
apply the admission test in
docs/product-scope.md, prefer consolidation, and record the measured cost, repeated workflow, or incorrect state. - Do not bump the version in feature or fix changes; add the changelog entry
under
## Unreleased. Only a requested release-preparation change bumps the version, after applying the cadence rules indocs/release.md#when-to-cut-a-release; if the pending changes do not meet them, recommend waiting. When your change meets that section's patch-release criteria, say so at handoff and recommend a prompt patch release. - For requested upgrades or mixed-version failures, read
docs/upgrading.md. Keep the target release, check the live daemon as well as CLI/MCP, and adapt affected consumers using the release migration. Do not silently downgrade to satisfy an old contract assertion or stale process; preserve queued work and the existing approval scope.
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.
- 9d ago Changed · +2 lines · +29 tokens per session 0f1412fa63c4
- 23d ago Changed · +5 lines · +73 tokens per session 56e0b7e740ea
- 26d ago Changed · +8 lines · +103 tokens per session d5fa3de50c1f
- 1mo ago Changed · +2 lines · +45 tokens per session 976f60c11a51
- 1mo ago Changed · +2 lines · +28 tokens per session 2f00d78a084e
- 1mo ago Changed · +10 lines · +177 tokens per session 2883ec60e094
- 1mo ago First seen · 75 lines · 796 tokens per session scan A 048ece328961
mergetrain AGENTS.md is an instructions file published in the GitHub repository yongjip/mergetrain (7 stars, last pushed today), licensed MIT. It adds 1,251 tokens to every session, about $0.0050 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.
Other instructions, from other repositories
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).
deno copilot-instructions.md
Copilot instructions for denoland/deno, covering deno development guide for github copilot, network access, high level overview, key directories and key files to understand first.