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/thixpin/pitway/agents-mdgit clone --depth 1 https://github.com/thixpin/pitwayWhat 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 | $0.01900 | $0.01900 |
| Opus 5 | $0.00950 | $0.00950 |
| Sonnet 5 | $0.00380 | $0.00380 |
| Haiku 4.5 | $0.00190 | $0.00190 |
Grade A, and why
pitway 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 — 73 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent Instructions
- This project uses PitWay to control the engineering workflow.
- Run
pitway resumebefore starting or resuming any work. - Never edit
.pitway/directly. - Work only within a confirmed task boundary.
- Obtain a task's bounded context via
pitway task-status <id> --context. - The driver protocol is split into roles:
protocol-driver.md(Main Agent: developer conversation and every approval gate),protocol-orchestrator.md(Orchestrator: task execution), andprotocol-worker.md(Worker: one bounded task) in the installed driver directory.
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
Project Status
PitWay is developed using its own workflow (dogfooding): every change lands through a confirmed milestone/contract/task, following the pipeline below. The authoritative sources are the workflow state in .pitway/ (milestones, contracts, backlog) — the original design record is archived at docs/archive/IMPLEMENTATION_PLAN-v1.0.md (frozen; never updated).
- Never write implementation code outside a confirmed milestone task. A milestone's contract must be presented to the developer and explicitly approved in conversation before its baseline commit or any task work.
- Check
.pitway/state.yamland the active milestone'scontract.md/tasks.yamlto find current state and next work before doing anything.
What PitWay Is
An npm-distributed CLI (npx pitway init; unscoped package name pitway) that controls the engineering process around AI coding agents — it is not itself an agent. Core philosophy: "Claude drives the interaction; PitWay controls the workflow state and engineering boundaries."
Pipeline: Requirement → Milestone → Contract → human confirmation → verification plan → right-sized task graph → task-level TDD (RED→GREEN→REFACTOR) → verification → atomic commit → milestone contract verification → completion.
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 · 73 lines · 1,900 tokens per session scan A 96a7c949fd0f
pitway AGENTS.md is an instructions file published in the GitHub repository thixpin/pitway (19 stars, last pushed 4d ago), licensed MIT. It adds 1,900 tokens to every session, about $0.0095 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
concord-mcp CLAUDE.md
Instructions for Get-Concord-AI/concord-mcp, covering concord mcp — engineering guide, what this project is, coding rules (non-negotiable), the "never typecast" pattern for db rows and external json and architecture.
paceflow CLAUDE.md
Instructions for paceaitian/paceflow, covering claude.md, 维护规则, 常用验证, 一条命令聚合全部核心套件 + claude plugin validate + git diff --check and 任一子套件非零退出即整体非零退出，杜绝手敲漏跑某套件致回归静默漏网.
agent-ready-repo AGENTS.md
Instructions for eugenelim/agent-ready-repo, covering agents.md, project overview, rule lookups, documentation and development workflow.
AgentRules AGENTS.md
Instructions for LearnAIHubC/AgentRules, covering agents.md, 1. think before coding, 2. simplicity first, 3. preserve existing architecture and framework boundaries and 4. surgical changes.
hardproof AGENTS.md
Instructions for asimons81/hardproof, covering hardproof agent instructions, project identity, read order, architecture map and release infrastructure.
lunar AGENTS.md
Instructions for gszr/lunar, covering how to work here, our goals, before you type, code and scope.