horosa-skill
Horace-Maxwell/horosa-skill/CLAUDE.md
Horosa Skill — Claude Code 入口 本仓把星阙(Horosa)的 114 个术数/占星技法打包成 local-first 的 MCP server + CLI(离线 runtime 走 GitHub Releases;MCP 全量面 127 = 114 技法 + 13 门面,`HOROSA
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.
Horace-Maxwell/horosa-skill/CLAUDE.md
Horosa Skill — Claude Code 入口 本仓把星阙(Horosa)的 114 个术数/占星技法打包成 local-first 的 MCP server + CLI(离线 runtime 走 GitHub Releases;MCP 全量面 127 = 114 技法 + 13 门面,`HOROSA
1c-syntax/bsl-language-server/CLAUDE.md
пользователя. То же касается вложенных `CLAUDE.md` в подкаталогах. ## Что это за проект **BSL Language Server** — реализация [LSP](https://microsoft.github.io/language-server-protocol/) для языка **1С:Предприятие 8 (BSL)** и [OneScript](http://oscript.io
humanlayer/agentcontrolplane/CLAUDE.md
entry points ├── config/ # Kubernetes manifests ├── internal/ │ ├── controller/ # Kubernetes controllers │ ├── llmclient/ # LLM provider clients │ ├── mcpmanager/ # MCP server management │ └── humanlayer/ # Human approval integration ├── docs/ # Comprehensive documentation └── test/ # Test suites ``` ## Common Commands
dromara/mica-mqtt/CLAUDE.md
Message processing pipeline - `MqttClientAioHandler` / `MqttClientAioListener`: t-io integration layer **Server Side:** - `MqttServer` / `MqttServerCreator`: Main server API with builder pattern - `MqttServerProcessor`: Message routing and delivery - `IMqttSessionManager`: Client session lifecycle management - `MqttMessageStore
nagisanzenin/skyclaw/CLAUDE.md
lab/witness/RESEARCH_PAPER.md for the theory and tems_lab/witness/EXPERIMENT_REPORT.md for empirical results. temm1e-mcp -- MCP client (stdio + HTTP, 14-server registry) temm1e-automation -- Cron jobs and scheduled tasks temm1e-observable -- OpenTelemetry tracing
theredsix/agent-browser-protocol/CLAUDE.md
console.log/warn/error, CORS errors, CSP violations, uncaught exceptions via WebContentsObserver (non-fingerprintable, no CDP) - **MCP Server**: Embedded MCP (JSON-RPC over HTTP) with 21 tools at `/mcp` ### Architecture ``` ┌─────────────────────────────────────────────┐ │ HTTP Client
daloopa/investing/CLAUDE.md
doing fundamental equity research. ## Analyst Context - Perspective: long/short equity, fundamental analysis - Data source: Daloopa MCP server (SEC filings, earnings, financial statements) - Market data: MCP-first (use whatever market data MCP
docfork/docfork/CLAUDE.md
dgrep build pnpm --filter dgrep typecheck pnpm --filter dgrep test pnpm --filter docfork build # MCP server package pnpm --filter docfork typecheck ``` ### Before committing ```bash pnpm lint:check && pnpm typecheck ``` ## Architecture
obra/episodic-memory/CLAUDE.md
esbuild`. Both are committed. - **CLI entry points** in `cli/` (Node.js wrappers — no bash). - **MCP server** is `dist/mcp-server.js`, launched by `cli/mcp-server-wrapper.js` from the plugin manifest. - **Tests** live in `test
juyterman1000/entroly/CLAUDE.md
Rust cd entroly-core && cargo clippy --all-targets -- -D warnings ``` ## Run ```bash entroly # Start MCP server (STDIO) entroly proxy # HTTP reverse proxy on localhost:9377 entroly go # Full onboarding (detect
Trevoke/org-gtd.el/CLAUDE.md
exports to Texinfo) - Comprehensive README.org with quick tour ## Memory Annotations - When using the emacs mcp server, always redefine or eval-buffer, don't use file loads, because of emacs' caching
fdmtl/director/CLAUDE.md
MCP gateway. It allows users to create and manage multiple playbooks. Each playbook aggregates MCP servers and prompts. We've just migrated director from a single tenant, local first architecture
angelnicolasc/graymatter/CLAUDE.md
answers. GrayMatter is a single-binary Go memory system for AI agents. Library + CLI + MCP server + TUI. Facts persist to bbolt with optional vector embeddings, for roughly
cyanheads/atlas-mcp-server/CLAUDE.md
stdio communication (default) - `httpTransport.ts` - HTTP server with authentication/rate limiting **MCP Layer** (`src/mcp/`): - `server.ts` - Main MCP server setup, tool/resource registration - `tools/` - MCP tool implementations (15 total tools) - `resources/` - MCP resource handlers
aaronsb/obsidian-mcp-plugin/CLAUDE.md
HTTP calls with direct Obsidian plugin API calls. This allows reuse of all existing MCP server logic while gaining performance benefits. ## Code Quality Guidelines ### SOLID Principles Application - **Single Responsibility**: - `ObsidianAPI
duolahypercho/fusion-fable/CLAUDE.md
# Claude Fable 5 — System Prompt --- Claude should never use {antml:voice_note} blocks, even
peakmojo/applescript-mcp/CLAUDE.md
Project Overview This is the AppleScript MCP (Model Context Protocol) Server - a dual-implementation MCP server that allows AI assistants to execute AppleScript on macOS systems. The server is available
zhizhuodemao/ai-reverse-toolkit/CLAUDE.md
populated). ## Key MCP Dependency This toolkit is designed to work with the **js-reverse MCP server** for browser debugging. The MCP provides Chrome DevTools Protocol access: network inspection, breakpoints, script
alibaba/UnifiedModel/CLAUDE.md
docs/en/guides/query-service.md` 5. `docs/en/reference/mcp.md` 6. `docs/en/architecture/overview.md` Important directories: - `cmd/umodel-server` - REST server. - `cmd/umctl` - public CLI. - `cmd/umodel-mcp` - MCP server. - `internal/query` - public read path implementation. - `internal/agentgateway` - agent-facing surface. - `internal/sampledata` - bundled sample loading
daydreamlive/scope/CLAUDE.md
requires local hardware access. ## Local Cloud Testing For local Livepeer cloud testing, follow `.agents/skills/testing-livepeer/SKILL.md`. ## MCP Server Testing When asked to test Scope via MCP tools (e.g., with a workflow JSON
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.