TaskFlow runs user-taught multi-step macros over existing TypeAgent actions. Flows are pure JSON — no TypeScript logic. A deterministic compile script handles all boilerplate generation (TypeScript type, grammar rule, manifest entry). When asked to "compile pending recipes" or similar: That's it. scripts/compileRecipes.mjs reads pending/*.recipe.json and: If the build fails, fix the error in the generated files and re-run pnpm run build. When asked to write a recipe for a new flow (or if you notice a missing flow): Do NOT…
Cupertino Panes is a TypeScript library that creates multi-functional modals, cards, and panes with touch technologies. It provides native iOS-style behaviors and animations for web applications. Each module can be enabled/disabled through the modules configuration option.
An MCP proxy server (Go, single main package, flat layout) that aggregates multiple downstream MCP servers behind one HTTP entrypoint. Each configured downstream server is exposed at its own route (<baseURL path>/<serverName>/) via SSE or streamable HTTP. make format expects golangci-lint and nilaway to be installed. Run it (or at least go vet + tests) before considering a change done. Request flow: MCP client → HTTP mux (http.go) → per-server middleware chain (recover, optional logging, optional bearer-token auth) → server.MCPServer…
This repository contains skills for AI agents working with dbt projects. This repo uses the superpowers skill framework. When creating or modifying skills: The superpowers marketplace is configured in .claude/settings.json and will be auto-installed when you trust this repo. Every SKILL.md must have valid frontmatter: Critical Rules: - name MUST be lowercase with hyphens only (letters, digits, hyphens) - name MUST match the directory name exactly - Only allowed fields: name, description, allowed-tools, compatibility, license, metadata, user-invocable - NO version,…
See AGENTS.md for development conventions. Quick reference: - Use typer.echo() for CLI output, logger from logging.py for progress - Run uv run ty check src/ before committing - Add tests for new features - Check dataclasses in models.py and runner.py when modifying commands
Memora is the sole memory system for this project. Do NOT use Claude Code's built-in auto memory (MEMORY.md). All memory operations must go through Memora MCP tools. Behavior is defined in .claude/skills/memora/SKILL.md.
This feeds the project entity compiler, which runs every 2 hours. If the save doesn't mention the project by name, it won't be compiled into the entity page.
All project guidance for coding agents lives in AGENTS.md (single source of truth for Claude Code, Codex, Cursor, Copilot and others). Do not add knowledge here; edit AGENTS.md.
@.context/INDEX.md This is a placeholder. Will need to be updated. https://github.com/kgoedecke/doop Doop is the open-source alternative to Paper.design: a multiplayer design canvas for humans and AI agents. Every design lives…
This repository packages 37signals' Rails patterns in two forms: There is no application code, build system, or tests. When editing or adding content: - Focus on transferable patterns, not Fizzy-specific…
- Write clear, concise commit messages - One logical change per commit - Run tests before committing - Update CHANGELOG.md in the same commit — every behaviour/config/docs/deps change ships a changelog entry (see rules/changelog.md). No commit without one. ALWAYS write tests before implementation. No exceptions. If asked to add a feature without tests, write the tests first then ask to proceed. If a PR diff shows untested code paths, flag them before continuing. Invoke: qwen3 "" Invoke: gemini --flash-lite ""
Follow the shared repository instructions in AGENTS.md. Agent workflows are in .agents/skills/; compatibility commands under .claude/commands/ invoke those workflows. Shared entrypoint for Claude, Codex, and humans. Keep this short and navigational; durable detail lives in docs/. - docs/index.md — system-of-record map. - docs/ARCHITECTURE.md — module map, the two-backend design, host→guest data flow, architectural invariants. - docs/trust-model.md — trust boundaries and taint sources (the security spec; read before touching secrets, subprocess, network, or the guest boundary). - docs/code-style.md — Rust authoring…
Read the shared development policy before repository work. It owns architecture, source ownership, verification, and delivery rules. This file owns host guidance only; keep common rules in that policy. Canonical spec scaffold: templates/spec.md. For skill/platform changes also read cross-platform patterns. On Claude Code, invoke skills by skill id, flow-next:flow-next-<name>; the /flow-next:<name> commands are typed by the user and hidden from the model. Other consumers of this file use their supported native equivalent. Distinguish Claude-native product source examples from the tools…
A file Claude Code reads at the start of every session. It holds the commands, conventions and warnings the agent needs for this project.
Where does it go?
At the repository root. Claude Code also reads CLAUDE.md files in subdirectories when it works there.
What should it contain?
Build and test commands, the project's layout, conventions that aren't obvious from the code, and mistakes to avoid. Short files tend to work better than long ones.
CLAUDE.md or AGENTS.md?
Claude Code reads CLAUDE.md; most other agents read AGENTS.md. Many projects keep one and point the other at it.
MicrosoftCLAUDE.md
