warp-memory-mcp
Provides read-only access to local Warp/Oz data, including saved memories, notebooks, conversations, user prompts, and project rules, and supports sidecar handoff notes for agent coordination.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@warp-memory-mcp Search Warp notebooks for scan_processing."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Warp Memory MCP Server
Read-only access to local Warp / Oz data so other agents (Grok, Claude, etc.) can coordinate with Warp’s AI.
Warp’s official Agent Memory lives on Oz (cloud, research preview, no public API yet). This server does not talk to that API. It reads what Warp stores on this Mac and a small sidecar used for Grok ↔ Oz handoffs.
What is available locally
Source | In | Notes |
Saved memories (AIFACT) |
| “Remember this” items Oz wrote locally |
Notebooks / plans |
| Full Warp Drive plan text |
Conversations |
| Title, first prompt, artifacts — not the transcript |
User prompts |
| The text you typed to Oz |
Project rules |
| Paths to |
Oz cloud Agent Memory | — | Not accessible |
Full chat transcripts | — | Cloud-only when Cloud Conversation Storage is on |
Related MCP server: artagon-vault-mcp
Prerequisites
uv is enough. No Full Disk Access is required —
the Warp DB lives in a Group Container, not ~/Library/Messages.
If Warp is running, the DB is opened read-only with a busy timeout.
Running
uv run --project /Users/rlittle/Development/warp-memory-mcp \
/Users/rlittle/Development/warp-memory-mcp/server.pyOverride the database path with WARP_SQLITE if needed.
Client config
Registered in:
~/.grok/config.tomlaswarp-memory(Grok)~/.warp/.mcp.jsonaswarp-memory(Oz can read Grok handoffs)
Restart Grok after adding the server so the tools appear.
Tools
Tool | Description |
| What is readable vs cloud-only, plus counts |
| Saved Warp AIFACT memories |
| Warp Drive notebooks / plans |
| Session summaries + linked plans + prompts |
| Recent user prompts to Oz |
| One search across memories, notebooks, conversations, prompts |
| Warp project rule files |
| Sidecar notes between Grok and Oz |
Handoffs are JSON files under ~/.warp/agent-coordination/handoffs/. They are
not written into warp.sqlite, so they will not fight Warp Drive sync.
Typical questions this answers
What did Oz prepare to work on tonight?
Search Warp notebooks for scan_processing.
Show the latest Oz conversation and its user prompts.
Write a handoff to Oz that Grok will do ScanImporter locking on 3.4beta.Safety
Warp’s database is read-only.
Command history (
commandstable) is not exposed.agent_tasksprotobuf blobs are not exposed.
Available Tools
13 toolsget_conversationB
Get one Warp conversation: summary, linked plan notebooks, and the user prompts from that session.
Inputs: id_or_conversation_id (required), query_limit.
get_handoffA
Read one handoff note by id.
Inputs: id (required).
get_memoryA
Get one Warp saved memory by numeric id or name.
Inputs: id_or_name (required).
get_notebookB
Get a Warp notebook / plan by id, title, or document uid.
Inputs: id_or_title (required).
list_conversationsB
List recent Warp agent conversations (title, first prompt, linked plans). Transcripts are not stored locally.
Inputs: limit, query, working_directory.
list_handoffsC
List Grok ↔ Oz coordination handoffs.
Inputs: limit, to_agent.
list_memoriesA
List Warp saved memories (local AIFACT 'remember this' items).
Inputs: query.
list_notebooksC
List Warp Drive notebooks / agent plans (full plans live here).
Inputs: query, include_body.
list_project_rulesA
List Warp project rule files (AGENTS.md / WARP.md paths).
Inputs: no named parameters.
list_queriesC
List recent user prompts sent to Warp / Oz.
Inputs: limit, query, conversation_id, working_directory.
searchA
Search Warp memories, notebooks/plans, conversation summaries, and recent user prompts to Oz. Best first tool for 'what did Oz work on' / 'find the Warp plan'.
Inputs: query (required), limit, sources.
statusA
Describe what Warp / Oz data is readable locally versus cloud-only. Use this first when coordinating with Warp's AI.
Inputs: no named parameters.
write_handoffC
Write a coordination note between Grok and Oz. Stored under ~/.warp/agent-coordination/handoffs (not in warp.sqlite).
Inputs: body (required), title (required), tags, to_agent, +3 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 13 tools
Each tool targets a distinct resource/action: status and search serve as meta/cross-cutting entry points, list_/get_ pairs cleanly separate memories, notebooks, conversations, and handoffs, and write_handoff is clearly the only write operation. Any conceptual overlap, such as search versus list_queries, is mitigated by clear descriptions.
Most tools follow a consistent list_<plural> / get_<singular> pattern, with write_handoff fitting the verb_noun convention. The exceptions are the bare commands status and search, which would be more consistent as get_status and search_warp, but the overall pattern remains readable and predictable.
Thirteen tools is well within the ideal scope and each tool has a clear role: four list/get resource pairs, a cross-resource search, a status overview, project rule listing, and a handoff write/list/get set. No tool feels redundant or superfluous.
The server provides strong read coverage for memories, notebooks, conversations, and recent prompts, plus a complete handoff write/list/get workflow. Minor gaps exist, such as project rules being listed only as paths without a content-read tool, but the core coordination and lookup workflows are well covered.
Maintenance
Related MCP Connectors
Agent-native MCP server over the public saagarpatel.dev corpus. Read-only, stateless.
Read-only MCP server exposing a user ORANO library to their own AI agent.
Guarded MCP server for agent-readable business truth, provenance, readiness, and discovery.
Read-only MCP server for the OPERANT AI operating-agent calibration benchmark.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceRead-only MCP server that exposes locally generated voice notes to any MCP client, allowing listing, searching, and reading transcripts, summaries, and knowledge graphs from voice-notes sessions.1MIT
- FlicenseNot gradedqualityCmaintenanceRead-only MCP server for querying an evidence-aware knowledge vault with temporal and provenance-aware data, supporting agent memory and semantic graph projections.-
- AlicenseNot gradedqualityCmaintenanceRead-only MCP server for TraceFact that lets agents list trace runs, fetch run details, and retrieve reliability reports from a local store.Apache 2.0
- AlicenseNot gradedqualityAmaintenanceRead-only local MCP server for inspecting AgentInspect TypeScript AI-agent traces. Lists runs, exposes bounded causal-failure and TraceFacts evidence, evaluates TraceContract checks, compares runs, and creates share-checked evidence without mutating traces or invoking agent tools.14,507 npm212MIT