agentleFS
yangjingoyangjingo/Synapse/AGENTS.md

Synapse

AGENTS.md4 starsChanged 3 months agoMITView on GitHub

SKILL.md is the entry point for the Synapse workflow. Specialized agent instructions live under skills/: synapse-forge shapes content, synapse-design defines visual rules, synapse-pretext renders editorial HTML, synapse-excalidraw creates diagrams, and synapse-animation handles motion exports. Each skill keeps its own SKILL.md, references, scripts, and examples where applicable. End-to-end sample outputs belong in examples/<topic>/; keep a topic's *-blog.dsl.md, *-blog.html, *-slides.dsl.md, *-slides.html, and figures/ together. Shared documentation and branding live in docs/ and assets/. There is no package manifest or compiled build. Work…

  • In this repositoryOne other file it gives its agents.

The file

Files
AGENTS.md

Repository Guidelines

Project Structure & Module Organization

SKILL.md is the entry point for the Synapse workflow. Specialized agent instructions live under skills/: synapse-forge shapes content, synapse-design defines visual rules, synapse-pretext renders editorial HTML, synapse-excalidraw creates diagrams, and synapse-animation handles motion exports. Each skill keeps its own SKILL.md, references, scripts, and examples where applicable. End-to-end sample outputs belong in examples/<topic>/; keep a topic's *-blog.dsl.md, *-blog.html, *-slides.dsl.md, *-slides.html, and figures/ together. Shared documentation and branding live in docs/ and assets/.

Build, Test, and Development Commands

There is no package manifest or compiled build. Work directly with the Markdown, HTML, JavaScript, and Python sources.

  • npx prettier --write path/to/file.html formats generated HTML.
  • python skills/synapse-forge/scripts/struct_validator.py examples/<topic> checks slide ordering and required structural sections.
  • python skills/synapse-forge/scripts/content_validator.py examples/<topic> checks cover, outline, takeaway, reference, and visual-content rules.
  • node skills/synapse-forge/scripts/url_validator.js --urls examples/<topic> verifies external links; this requires network access.
  • node skills/synapse-excalidraw/scripts/validate.js --strict path/to/file.excalidraw detects diagram collisions and arrow crossings.

Open generated HTML in a browser to confirm typography, overflow, images, and slide navigation.

Coding Style & Naming Conventions

Follow the surrounding file: JavaScript uses two-space indentation, semicolons, single quotes, and camelCase; Python uses four spaces, snake_case, type hints where useful, and standard-library imports first. Keep Markdown headings short and imperative. Use lowercase kebab-case for topic directories and artifact stems, such as deepseek-v4/deepseek-v4-slides.html. Name figures sequentially (figure-01-cost-curve.svg) and keep editable .excalidraw sources beside exports.

Testing Guidelines

No centralized test runner or coverage threshold exists. Run the relevant validators for every changed artifact and manually inspect rendered output. When changing a validator, exercise both a known-good example and a deliberately invalid temporary fixture. Avoid committing browser screenshots or generated debug output unless they are intentional example assets.

Commit & Pull Request Guidelines

History follows Conventional Commit prefixes, chiefly feat:, docs:, and chore:. Write a concise imperative subject describing one logical change. Pull requests should summarize affected skills/examples, list validation commands run, link relevant issues or source material, and include before/after screenshots for visual changes. Do not mix unrelated regenerated artifacts into the same PR.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.