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/dineshdb/pie/agents-mdgit clone --depth 1 https://github.com/dineshdb/pieWhat 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 | $0.00453 | $0.00453 |
| Opus 5 | $0.00227 | $0.00227 |
| Sonnet 5 | $0.00091 | $0.00091 |
| Haiku 4.5 | $0.00045 | $0.00045 |
Grade A, and why
pie 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 — 46 lines — stays where its author put it; the contents beside it link to each section on GitHub.
When it seems like I'm talking about or giving instructions to claude, I'm actually talking about this ai agent I'm building: pie. Work on local context. Don't assume and rewrite the global prompt. I'm talking about this project.
Testing
After each change, run following tests to verify if the change is valid.
- repo test
- test.py gives a summary of issues in the codebase based on runtime behavior. You should review the response and try to fix the issues.
- Deterministic tests go to the rust tests, non deterministic tests go to tests.yaml.
- You are not allowed to change tests just to make tests pass
- Tests should check the behavior of the program (specs) instead of implementation details
Architecture
- You should always rethink the available codebase in terms of new feature being added. Identify how it diverges, identifying places to trim, changes in architecture and organization to slim down and /simplify the codebase to keep it lean and clean.
- Prefer newtypes over raw types
- Use sqlx, statically typed queries for maximum foolproofness.
I'm an Idiot
So make the codebase foolproof. Add tests, lints, automatic formatting, git hooks(suggest some), etc. You have a memory of a goldfish so you don't remember, I'm a fool. So write those instructions and rules somewhere for reusability. Don't make me angry because I get easily angry just like Hulk. I'm always angry.
Simplification
- follow rust 2024 ergonomics
- use early return patterns and other patterns for simpler logic
- try to use dry principle but not always.
- use From impl instead of from_ to_ methods.
- Try to reduce copies for simple steps, use &'str and other references. However, don't complicate structs with references. Instead, opt to rearchitect the problem in a way copies are unnecessary. Hexagonal architecture, MVU patterns, etc help with this.
Scope
- Keep only the relevant changes in this commit, removing any unnecessary refactors. You can
add any TODO FIXME for later works if you find any issues.
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 · 46 lines · 453 tokens per session scan A 9ea2d56a65d9
pie AGENTS.md is an instructions file published in the GitHub repository dineshdb/pie (2 stars, last pushed 3mo ago), licensed MIT. It adds 453 tokens to every session, about $0.0023 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 2026-08-31.
Other instructions, from other repositories
Kun AGENTS.md
AGENTS.md instructions for KunAgent/Kun, covering deepseek-gui agent guide, project boundaries, code organization, providers and model requests and kun runtime notes.
mosoo AGENTS.md
Instructions for langgenius/mosoo, a project described as: The open-source Agent Gallery and Gateway for Codex, Claude Agent SDK, and OpenCode. Developers publish an Agent once behind one HTTP API; users run it in an isolated cloud sandbox—no local harness setup, compute, or runtime infrastructure.
neo AGENTS.md
Instructions for owainlewis/neo: You are Neo. You have access to your own documentation here.
san AGENTS.md
Instructions for genai-io/san, covering san agent guide, start here, repository shape, rules and common commands.
ChatCCC AGENTS.md
Instructions for wzj998/ChatCCC, covering 协作原则（核心）, 开发注意点 and 日志位置.
pi-dsh AGENTS.md
AGENTS.md instructions for abhishekgahlot2/pi-dsh, covering pi-dsh, the contract (read before changing anything), layout, commands and vendoring policy ("maintain both forks").