# ContextStream

Unify workspace memory, project context, and code intelligence to deliver the right details instantly. Index repos, track decisions, and explore dependency relationships to speed up onboarding and ch…

## Quick Start

```bash
# Connect this server (installs CLI if needed)
npx -y @smithery/cli@latest mcp add contextstream/mcp-server

# Browse available tools
npx -y @smithery/cli@latest tool list contextstream/mcp-server

# Get full schema for a tool
npx -y @smithery/cli@latest tool get contextstream/mcp-server mcp_server_version

# Call a tool
npx -y @smithery/cli@latest tool call contextstream/mcp-server mcp_server_version '{}'
```

## Direct MCP Connection

Endpoint: `https://mcp-server--contextstream.run.tools`

**Optional config:**
- `apiKey` (query) — Optional ContextStream API key or JWT (required for authenticated tool calls).
- `projectId` (query) — Optional project ID to scope requests.
- `workspaceId` (query) — Optional workspace ID to scope requests.

## Tools (59)

- `mcp_server_version` — Return the running ContextStream MCP server package version
- `auth_me` — Fetch authenticated user profile
- `workspaces_list` — List accessible workspaces (paginated list: items, total, page, per_page, has_next, has_prev).
- `projects_list` — List projects (optionally by workspace; paginated list: items, total, page, per_page, has_next, has_prev).
- `projects_create` — Create a new project within a workspace.
- `projects_update` — Update a project (rename or change description)
- `projects_index` — Trigger indexing for a project
- `search_semantic` — Semantic vector search
- `search_hybrid` — Hybrid search (semantic + keyword)
- `search_keyword` — Keyword search
- `memory_create_event` — Create a memory event for a workspace/project
- `memory_list_events` — List memory events (optionally scoped)
- `memory_create_node` — Create a knowledge node with optional relations
- `memory_list_nodes` — List knowledge graph nodes
- `memory_search` — Search memory events/notes
- `memory_decisions` — List decision summaries from workspace memory
- `graph_related` — Find related nodes in the knowledge graph
- `graph_path` — Find path between two nodes
- `graph_decisions` — Decision history in the knowledge graph
- `graph_dependencies` — Dependency graph query
- `graph_call_path` — Find call path between two targets
- `graph_impact` — Analyze impact of a target node
- `graph_ingest` — Build and persist the dependency graph for a project. Runs async by default (wait=false) and can take a few minutes for…
- `projects_get` — Get project details by ID
- `projects_overview` — Get project overview with summary information
- `projects_statistics` — Get project statistics (files, lines, complexity)
- `projects_files` — List all indexed files in a project
- `projects_index_status` — Get project indexing status
- `projects_ingest_local` — Read ALL files from a local directory and ingest them for indexing.
- `workspaces_get` — Get workspace details by ID
- `memory_get_event` — Get a specific memory event by ID with FULL content (not truncated). Use this when you need the complete content of a m…
- `memory_update_event` — Update a memory event
- `memory_delete_event` — Delete a memory event
- `memory_timeline` — Get chronological timeline of memory events for a workspace
- `memory_summary` — Get condensed summary of workspace memory
- `graph_circular_dependencies` — Detect circular dependencies in project code
- `graph_unused_code` — Detect unused code in project
- `session_init` — Initialize a new conversation session and automatically retrieve relevant context.
- `session_tools` — Get an ultra-compact list of all available ContextStream MCP tools.
- `session_get_user_context` — Retrieve user preferences, coding style, and persona from memory.
- `workspace_associate` — Associate a folder/repo with a workspace after user selection.
- `workspace_bootstrap` — Create a new workspace (user-provided name) and onboard the current folder as a project.
- `session_capture` — Automatically capture and store important context from the conversation.
- `session_capture_lesson` — Capture a lesson learned from a mistake or correction.
- `session_get_lessons` — Retrieve lessons learned from past mistakes and corrections.
- `session_smart_search` — Search memory with automatic context enrichment.
- `session_remember` — Quick way to store something in memory. Use natural language.
- `session_recall` — Quick way to recall relevant context. Use natural language.
- `generate_editor_rules` — Generate AI rule files for editors (Windsurf, Cursor, Cline, Kilo Code, Roo Code, Claude Code, Aider).
- `session_summary` — Get a compact, token-efficient summary of workspace context (~500 tokens).
- `session_compress` — Extract and store key information from chat history as memory events.
- `session_delta` — Get new context added since a specific timestamp.
- `context_smart` — **CALL THIS BEFORE EVERY AI RESPONSE** to get relevant context.
- `reminders_list` — List all reminders for the current user.
- `reminders_active` — Get active reminders that are pending, overdue, or due soon.
- `reminders_create` — Create a new reminder for a specific date/time.
- `reminders_snooze` — Snooze a reminder until a later time.
- `reminders_complete` — Mark a reminder as completed.
- `reminders_dismiss` — Dismiss a reminder without completing it.

```bash
# Get full input/output schema for a tool
npx -y @smithery/cli@latest tool get contextstream/mcp-server <tool-name>
```

## Resources

- `contextstream:openapi` — Machine-readable OpenAPI from the configured API endpoint
- `contextstream:workspaces` — List of accessible workspaces

## Prompts (22)

- `explore-codebase` — Get an overview of a project codebase structure and key components
- `capture-decision` — Document an architectural or technical decision in workspace memory
- `review-context` — Build context for reviewing code changes
- `investigate-bug` — Build context for debugging an issue
- `explore-knowledge` — Navigate and understand the knowledge graph for a workspace
- `onboard-to-project` — Generate onboarding context for a new team member
- `analyze-refactoring` — Analyze a codebase for refactoring opportunities
- `build-context` — Build comprehensive context for an LLM task
- `smart-search` — Search across memory, decisions, and code for a query
- `recall-context` — Retrieve relevant past decisions and memory for a query
- `session-summary` — Get a compact summary of workspace/project context
- `capture-lesson` — Record a lesson learned from an error or correction
- `capture-preference` — Save a user preference to memory
- `capture-task` — Capture an action item into memory
- `capture-bug` — Capture a bug report into workspace memory
- `capture-feature` — Capture a feature request into workspace memory
- `generate-plan` — Generate a development plan from a description
- `generate-tasks` — Generate actionable tasks from a plan or description
- `token-budget-context` — Get the most relevant context that fits within a token budget
- `find-todos` — Scan the codebase for TODO/FIXME/HACK notes and summarize
- `generate-editor-rules` — Generate ContextStream AI rule files for your editor
- `index-local-repo` — Ingest local files into ContextStream for indexing/search