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/m-pineapple/warp-command-runner/agents-mdgit clone --depth 1 https://github.com/M-Pineapple/warp-command-runnerWhat 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 | $0.00645 | $0.00645 |
| Opus 5 | $0.00322 | $0.00322 |
| Sonnet 5 | $0.00129 | $0.00129 |
| Haiku 4.5 | $0.00064 | $0.00064 |
Grade A, and why
warp-command-runner 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 2d 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 — 50 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Warp Command Runner
Public MIT MCP server (Pineapple / M-Pineapple). Speaks MCP over stdio by default. Opt-in --http serves Streamable HTTP on 127.0.0.1 only. Users bring their own HTTPS tunnel. This project does not host a relay.
Done means the verification gate is green, docs match the binary, and the tree has no private identity.
Boundaries
- Public identity is Pineapple / M-Pineapple only. No private-estate names, maintainer tunnels, or real hostnames in git. Sample origin:
https://mcp.example.com. --httpbinds loopback only. Never0.0.0.0.- Stdio owns stdout. Logs go to stderr.
- Do not commit tokens, tunnel credentials, or
oauth-store.json. - No money-path. This is not a trading product.
Coding standards
SwiftPM CLI + MCP. Match existing files. ArgumentParser for flags. Agent workflow when driving this server: docs/AGENT_CODING_SOP.md.
Verification gate
Run from the repo root on macOS. Quote the summary lines. Triage every failure. There is no Xcode project; skip xcodebuild.
swift package resolve && ./scripts/patch-swift-sdk.sh && swift build && swift testGreen:Test Suite 'All tests' passedwith 0 failures.swift build -c releaseGreen:.build/release/warp-command-runnerexists.- Stdio MCP proof against that binary (not debug):
./examples/run_test.sh .build/release/warp-command-runnerGreen:initialize → Warp Command Runner v<version>andtools/list → 40 tools. - Remote HTTP proof against the same binary:
./examples/run_http_test.sh .build/release/warp-command-runnerGreen: lineLOOPBACK_OKand initialize version matchingAppIdentity.version. - Version:
AppIdentity.version,.build/release/warp-command-runner --version, CHANGELOG[X.Y.Z], and README "What's new in vX.Y.Z" agree.
Known failures: none. ProcessRunnerTests.testTimeoutTerminatesHungProcess waits on a real hung child (timeout API), not a test sleep.
Whole-tree hazard: scripts/patch-swift-sdk.sh is required while pinned to swift-sdk 0.10.x. swift test without the patch fails to compile the SDK.
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.
- 2d ago First seen · 50 lines · 645 tokens per session scan A 62118e3bac14
warp-command-runner AGENTS.md is an instructions file published in the GitHub repository M-Pineapple/warp-command-runner (17 stars, last pushed 5d ago), licensed MIT. It adds 645 tokens to every session, about $0.0032 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 2026-08-30.
Other instructions, from other repositories
happy AGENTS.md
Instructions for slopus/happy, covering agent workflow and sync to main.
the-unofficial-swift-concurrency-migration-skill CLAUDE.md
Instructions for kylehughes/the-unofficial-swift-concurrency-migration-skill, covering claude.md, repository purpose, build commands, generate the skill package and build options.
study-ios AGENTS.md
Instructions for ITlearning/study-ios, covering study-ios, skills, project structure and conventions.
ASD-AuDHD-PAI-Skills CLAUDE.md
Instructions for emory/ASD-AuDHD-PAI-Skills: You are the personal AI infrastructure of your user and this repository contains some Claude Skills that are intended to be used by someone that cares for or works with people with so-called "AuDHD" (ADHD+ASD), or other manifestation of pathological demand avoidance.
the-unofficial-swift-programming-language-skill CLAUDE.md
Instructions for Shada01245/the-unofficial-swift-programming-language-skill, covering claude.md, repository purpose, build commands, generate the skill package and build options.
the-unofficial-swift-programming-language-skill AGENTS.md
Instructions for Shada01245/the-unofficial-swift-programming-language-skill, a project described as: 📦 Unlock the Swift Programming Language with this structured Skill for LLMs, providing complete access to essential Swift content.