Skip to content

CONTRIBUTING.md

Latest commit

 

History

History
53 lines (36 loc) · 1.7 KB

CONTRIBUTING.md

File metadata and controls

53 lines (36 loc) · 1.7 KB
Raw

Contributing

Setup

1. Clone the repository

git clone https://github.com/M4SS-Code/pi-packages.git
cd pi-packages

2. Install dependencies

pnpm install

3. Install into pi (local)

Use pi install with a local path to register each extension for development:

pi install ./packages/binary-file-guard
pi install ./packages/llms-txt
pi install ./packages/search-delegator
pi install ./packages/staan-search
pi install ./packages/web-fetch

Development

pnpm run check:all    # type-check all packages
pnpm run format       # format all files

Issues and Pull Requests

You are responsible for what you submit. We don't care whether you wrote it yourself or used an AI; we care that it's good.

Every issue and PR must start with a short Intent section that clearly answers:

  • What problem are you solving?
  • Why does it matter?
  • Who is it for?

Example:

Intent: When the LLM tries to read a PDF, SQLite database, or archive, it dumps binary garbage into context instead of getting useful text. I want an extension that intercepts the read call, detects problematic files, and tells the model to use bash with the right CLI tool instead. This affects every user who works with non-text files in their projects.

I'm proposing a package rather than a skill because a skill adds overhead to the model's context and relies on the model remembering to read it. A package runs automatically on every read call with zero context cost.

Don't lead with the "how." Put the implementation details in the body; the first thing we need to know is why the change exists. If the intent isn't clear, we may close the issue/PR and may ban the user.