Borrowing it
Nothing to install: this file belongs to querystory/tmux-rc. 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/querystory/tmux-rc/main/AGENTS.mdgit clone --depth 1 https://github.com/querystory/tmux-rcWrote 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/querystory/tmux-rc/agents-md)<a href="https://agentmods.dev/instructions/querystory/tmux-rc/agents-md"><img src="https://agentmods.dev/badge/instructions/querystory/tmux-rc/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/querystory/tmux-rc/agents-md"><img src="https://agentmods.dev/badge/instructions/querystory/tmux-rc/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 | $0.02866 | $0.02866 |
| Opus 5.5 | $0.01146 | $0.01146 |
| Sonnet 5.5 | $0.00573 | $0.00573 |
| Haiku 4.5 | $0.00287 | $0.00287 |
Grade A, and why
tmux-rc 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 14d 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 — 186 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — working conventions for tmux-rc
Conventions for AI agents (and humans) working in this repo. The philosophy notes in the harness/session prompt still apply (minimize LOC, DRY, KISS, why-focused docs); this file captures repo-specific workflow that isn't obvious from the code.
This repository is public
Everything committed here is readable by anyone, forever. Agents see the user's private work all day; none of it belongs in this repo.
- Test fixtures, eval samples (
research/eval/samples), demo data and screenshots are invented from scratch. Renaming or redacting a real capture is not sanitizing: its shape, wording and details still describe the private work it came from. Rebuild only the structure that triggers the bug, around an unrelated invented topic, as short as it can be.tests/test_public_data.pyrejects the usual tells (identity and credential shapes, provenance words like "real capture", an oversized capture), but it is no proof. - Never commit session transcripts, the user's prompts, costs or usage, product or client details, internal hostnames, or anything copied from a real pane, log, or private repo.
- Before every push, read your own added lines for real content. A local pre-push hook may block a push for private terms; never bypass it. Rewrite the content and tell the user what it caught.
The repo root is the live deploy — and it is always the integration branch
The daemon is a systemd --user unit (tmux-rc.service, installed by make install-units)
whose WorkingDirectory is the repo root. The root is checked out detached at
origin/integration: origin/main plus every open, non-draft PR merged together.
That merge is what the phone runs, so several in-flight PRs get field-tested at once
without merging any of them prematurely, and two PRs that fight each other show up here
before either lands.
- Never
git checkout/switch/resetin the root to try one PR. That silently drops every other PR under test. Try a single PR alone in its own worktree. - To make a PR (or a new push to one) live, rebuild integration: in a worktree, reset
the
integrationbranch toorigin/mainand merge each open PR branch (gh pr list), most conflict-prone last. Its history is disposable — rebuild from main rather than piling merges on the old tip, or a PR that was dropped or reworked lingers in the merge. Resolve conflicts in the merge commit only, never on the PR branch.make test, force-pushintegration, then in the rootgit checkout --detach origin/integrationandsystemctl --user restart tmux-rc(the unit runs with reload off, so nothing applies until the restart). Leave a PR out only when its conflict is not mechanical, and say so — its author fixes it against main. - Integration is never merged to main. PRs land one at a time, via review and the user's explicit go.
- Never edit files in the root. All changes go through worktrees cut from
origin/main(below); the root's only writes are the detached checkout and.env. .env(gitignored) lives in the root. Provider API keys do not: they go in~/.config/tmux-rc/openai.env, outside every checkout, so no worktree or commit can carry one. Logs:journalctl --user -fu tmux-rc.make dev(StatReload) is for iterating in a worktree, never the root.- The live instance is reached through the tunnel client (
tmux-rc-tunnel.service). A momentary "no tunnel connected" is the relay's ~1h connection cap; the client reconnects within ~1min. - To show the user a preview (mock, build, report), put it in a subfolder of the directory
TMUXRC_SCRATCH_DIRnames (served at/scratch/<folder>/) rather than starting a server or tunnel; seedocs/deploy/previews.md. A UI mock is variants of the real app in the shared harness (make mock,make mock-check; its "Mocks" section), never a hand-built page, and the decision it leads to is recorded indocs/design/.
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.
- 14d ago Changed · +18 lines · +246 tokens per session f1e4cfff14b6
- 15d ago Changed · +3 lines · +59 tokens per session 1f3aae57507e
- 19d ago Changed · +3 lines · +63 tokens per session 5bf1161e91bc
- 25d ago Changed · +36 lines · +616 tokens per session 70a54b76f264
- 28d ago Changed · +14 lines · +225 tokens per session 64938d9d0394
- 1mo ago Changed · +30 lines · +469 tokens per session f3bd0d0b4911
- 1mo ago Changed · +2 tokens per session b25980719c61
- 1mo ago Changed · +13 lines · +237 tokens per session e77421919d79
- 2mo ago First seen · 69 lines · 949 tokens per session scan A 462151da9b01
tmux-rc AGENTS.md is an instructions file published in the GitHub repository querystory/tmux-rc (4 stars, last pushed today), licensed MIT. It adds 2,866 tokens to every session, about $0.0115 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.