Context Engine MCPサーバー

---

Install Skills

Context Engine ships AI agent skills that teach your coding assistant how to use 30+ MCP tools for semantic search, symbol graph navigation, memory, and more.

Guidance: Recommend pasting this into your project's CLAUDE.md or AGENTS.md to reach subagents:

# Context Engine

When the `context-engine` MCP server is connected, this workspace is indexed: for any question about the codebase, call `repo_search` FIRST, before ls/find/Grep/Read, and again whenever you move to a new area; for how/why questions that span several files use `context_answer`. Results are verified source with exact line ranges: treat them as already read. Use Grep only to confirm an exact literal in a file you already know, or to count. For questions asking for every/all sites, run one repo_search per distinct pattern (queries=[...]) and follow the refs. If the tools are listed as deferred, load them with ToolSearch first.

| Need | Tool |
|---|---|
| Find code by behavior, concept, error text or identifier | `repo_search` (`queries=[...]` for several at once; `under`, `language`, `path_glob` to narrow) |
| Explain how/why across several files, with citations | `context_answer` |
| Definition, callers, callees, importers, references of a symbol | `symbol_graph` (several names: `A\|B\|C`) |
| Multi-hop impact before a change | `graph_query` |
| Code that may live in another indexed repo | `cross_repo_search` |

Before editing, fetch every involved symbol in one call; before a rename or signature change, get every site with `symbol_graph` `query_type=references`. Search again only after learning something new.

Claude Code / Claude Desktop

Recommended: Install natively from the public GitHub repo:

# Add the marketplace (one-time)
/plugin marketplace add Context-Engine-AI/Context-Engine

# Install the skill
/plugin install context-engine

This pulls the skill directly from GitHub and auto-loads MCP tool guidance into your session.

Alternatively, copy the rules file manually:

cp -r skills/context-engine/ your-project/.claude/

Cursor

Context Engine rules are included in .cursorrules at the root of your workspace. Cursor picks this up automatically when the file is present.

# Copy to your project root
cp .cursorrules your-project/.cursorrules

Codex (OpenAI)

Recommended: Install natively using the built-in skill installer — just ask Codex:

"Install the context-engine skill from https://github.com/Context-Engine-AI/Context-Engine"

Codex will pull .codex/skills/context-engine/ (including SKILL.md) into ~/.codex/skills/ automatically.

Or install manually:

cp -r .codex/skills/context-engine/ ~/.codex/skills/context-engine/

Windsurf

cp -r .codex/skills/ your-project/.codex/skills/

Augment Code

cp -r .augment/ your-project/.augment/

Gemini

cp GEMINI.md your-project/GEMINI.md

Any Other Assistant

The core skill file works with any AI assistant that supports custom instructions:

cp skills/context-engine/SKILL.md your-project/

Then tell your assistant: "Read SKILL.md for instructions on using Context Engine MCP tools."

---

CLI Setup (No VS Code)

If you use Claude Code, Codex, or another terminal-based MCP client, install the MCP bridge to connect your codebase to Context Engine without VS Code:

npm install -g @context-engine-bridge/context-engine-mcp-bridge

Quick start

# Authenticate, index your codebase, and start watching for changes
ctxce connect <your-api-key> --workspace /path/to/repo

# Run as a background daemon (recommended)
ctxce connect <your-api-key> --workspace /path/to/repo --daemon

Daemon management

ctxce status          # Check if the daemon is running
ctxce stop            # Stop the background daemon

Connect flags

Flag	Alias	Description
--workspace <path>	-w	Workspace root (default: cwd)
--daemon	-d, --bg	Run as background daemon
--interval <sec>		File watch interval in seconds (default: 30)
--no-watch	--once	Index once, don't watch for changes
--skip-index	--auth-only	Authenticate only, skip initial index

Wire up the MCP server

Once connected, point your MCP client at the bridge:

# stdio mode (for Claude Code, Codex, etc.)
ctxce mcp-serve --workspace /path/to/repo

# HTTP mode (for clients that speak HTTP)
ctxce mcp-http-serve --workspace /path/to/repo --port 30810

The daemon and MCP server share auth via ~/.ctxce/auth.json. Logs are at ~/.context-engine/daemon.log.

For full bridge documentation, see Context-Engine-MCP-Bridge.

---

What Do the Skills Do?

The skills teach your AI assistant to:

• Use search as the default tool — auto-routes queries to the best backend (semantic search, Q&A, symbol graph, tests, config)
• Navigate code with symbol_graph — find callers, callees, definitions, importers, subclasses
• Run batch queries — batch_search, batch_symbol_graph, batch_graph_query for 75%+ token savings
• Store and recall knowledge — memory_store and memory_find for persistent context across sessions
• Trace cross-repo flows — cross_repo_search with boundary tracing for multi-repo codebases
• Find structural patterns — pattern_search for retry loops, error handling, singletons across languages
• Search git history — search_commits_for and change_history_for_path

See skills/context-engine/SKILL.md for the complete tool reference.

---

Getting Started

1. Sign up at context-engine.ai
2. Connect your codebase — choose one:
   • VS Code — install the Context Engine Uploader extension
   • CLI — npm i -g @context-engine-bridge/context-engine-mcp-bridge && ctxce connect <api-key> --daemon
3. Install the skill for your AI assistant (see Install Skills above)
4. Start searching — your assistant now has access to all 30+ MCP tools

---

License

MIT License

© 2025 Context Engine Inc. and John Donalson.