How real projects brief Claude Code: the commands, conventions and warnings they put in CLAUDE.md.
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.
PostHogReads credentialsCLAUDE.mdPostHog/feature-flag-demo/CLAUDE.md
Feature flag demonstration application built with Next.js 16, React 19, and PostHog integration. The application demonstrates real-time server-side logging with SSE (Server-Sent Events) and feature flag evaluation patterns. Important: This…
PostHogReads credentialsCLAUDE.mdPostHog/finops-demo/CLAUDE.md
This file provides guidance to Claude Code when working with this repository. FinOps Detective — a demo app for "FinOps as MCPs in a Trench Coat." Shows how to use…
PostHogCLAUDE.mdPostHog/HogTTV/CLAUDE.md
A browser extension for Chrome and Firefox that renders custom Slack emojis in Google Meet chat. One manifest.json serves both browsers.
PostHogCLAUDE.mdPostHog/kea-disposables/CLAUDE.md
kea-disposables — a Kea plugin that gives every mounted logic a cache.disposables manager for useEffect-style resource cleanup, plus automatic pause/resume of that work while the browser tab is hidden. Extracted from frontend/src/kea-disposables.ts in PostHog/posthog, which remains its largest consumer. - One DisposablesManager per logic, stored on logic.cache.disposables in the plugin's afterMount, torn down in beforeUnmount — but only when logic.isMounted() has gone false, because kea reference-counts mounts. - Every manager is also registered in a module-level globalVisibilityState.allManagers set. A single…
PostHogCLAUDE.mdPostHog/libdebugger/CLAUDE.md
libdebugger is a Python library that implements runtime bytecode instrumentation for live debugging. It integrates with PostHog to enable dynamic probes that can be installed remotely and capture function entry/exit…
PostHogReads credentialsCLAUDE.mdPostHog/posthog-wii/CLAUDE.md
Reads the play history that the Wii System Menu keeps in its Message Board file and sends it to PostHog as events. Two front ends share one C core: a…
PostHogCLAUDE.mdPostHog/postpile/apps/desktop/CLAUDE.md
Read these before editing anything under src/renderer/src/. The visual reference is the "Crisp native, refined" mockup (StyleCrispPro) with the "Warm reach" tile look on top (see "Tile look" below); the layout is still open, so keep components small and cheap to move. The renderer shows what the local API returned and nothing else. - Sample data lives in one place: FakeEngine (apps/server), switched on with POSTPILE_FAKE=1. The renderer never has its own fixtures and never has a fake ? sample…
PostHogCLAUDE.mdPostHog/splunk-add-on/CLAUDE.md
A Splunk add-on that polls the PostHog activity logs API and indexes each entry as an OCSF event. It exists because Splunk Cloud does not let customers create scripted inputs, and the roles available to them do not include the capability to add one, so polling from Splunk Cloud requires a packaged add-on. Change globalConfig.json rather than anything under output/. UCC regenerates that directory on every build, so edits there vanish. The version lives in one place: meta.version in globalConfig.json.
PydanticCLAUDE.mdpydantic/demo-endpoints/CLAUDE.md
This is a Cloudflare Workers project that serves random content through REST API endpoints. The project follows a simple Cloudflare Workers architecture: The main fetch handler uses URL pathname matching to route requests to different endpoints. Each endpoint contains its own logic for generating random responses. The project does not have tests.
PydanticCLAUDE.mdpydantic/pydantic-who/CLAUDE.md
ALWAYS run make lint after editing code to format and check linting/typechecking.
A file Claude Code reads at the start of every session. It holds the commands, conventions and warnings the agent needs for this project.
At the repository root. Claude Code also reads CLAUDE.md files in subdirectories when it works there.
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 Code reads CLAUDE.md; most other agents read AGENTS.md. Many projects keep one and point the other at it.
