1. Conduid
  2. RAG
  3. Papez
MCP server · RAG

Papez

Open-source causal graph memory for AI agents. 85.55 on LoCoMo under a frozen, reproducible protocol. MCP server with ACT-R scoring, spreading activation, and active forgetting.

58Fair

Scored just now · breakdown

About Papez

Papez is an MCP server published by Astrix-Labs in the RAG category: open-source causal graph memory for AI agents. 85.55 on LoCoMo under a frozen, reproducible protocol. MCP server with ACT-R scoring, spreading activation, and active forgetting.

The repository has 31 stars and 5 forks, with the last commit 4 days ago. Recent commit activity is the strongest signal that reported issues will get attention.

Install

uvx
uvx papez
pip
pip install papez
From source
git clone https://github.com/Astrix-Labs/papez

This server has no ConduID identity, so agent calls to it are not receipted. Pin the version you install and review the source before granting it credentials.

Tools (24)

delete_memoryPermanently delete a memory (writer+ for space memories)

Parameters not published.

list_core_memoriesList core memories, optionally filtered by category. Optional `space`.

Parameters not published.

memory_amendRecord a correction: creates a new memory that **supersedes** an existing one. The old memory is kept (decayed in recall), not deleted. Optional `refs` / `tags` (omitted: copied from the old memory; given: replace). Stays in the old memory's space.

Parameters not published.

memory_explainExplain a memory's score. Includes a `score_model` block (formula + live per-force breakdown + staleness note), `removal_impact`, and the memory's `refs`, `tags`, `space`.

Parameters not published.

memory_recallRecall memories by natural language query (vector + keyword + graph spreading activation). Supports `verbosity: "concise"` or `"light"` for lightweight payloads. Optional `tags_any`, `ref_type`, `space`, `spaces`, `include_personal`, `category`, `exclude_superseded`.

Parameters not published.

memory_retire_by_refRetire the memories citing a removed source (`ref_type`, `ref_id`). `mode="amend"` (default) or `"delete"`. Optional `space`.

Parameters not published.

memory_searchFiltered vector search by status, category, date (`since`), last-active date (`active_since`), entity, and now `tags_any`, `tags_all`, `ref_type`, `ref_id`. Pass an **empty query** to enumerate by recency instead (no embedder needed) — with `since`/`active_since` this answers "what's new since I last looked" without knowing what to query for. Optional `space`, `spaces`, `include_personal`.

Parameters not published.

memory_statsGet memory system statistics over what you can see. Optional `space`.

Parameters not published.

memory_storeStore a new memory. Use `related` for writer-specified **typed** edges (`{id, type}`); `related_to` is legacy and always creates `caused_by`. Optional `category`, `refs`, `tags`, `space`, `external_id`, `on_existing` (`amend` default, `skip`, `error`). May return `possible_conflicts` (heuristic hints).

Parameters not published.

memory_store_batchStore up to 100 memories into one space in one call. Validates every item first (all or nothing), supports `ref_index` links to earlier items, returns one result per item. Safe to retry when every item has an `external_id`.

Parameters not published.

Show all 24 tools
memory_traverseWalk the causal graph from a node. Returns reachable **nodes and the edges** of the induced subgraph (`source/target/type/weight/created_by`) — a superset of the BFS tree, so paths can be reconstructed. Honors `edge_types`. Optional `space`, `spaces`, `include_personal`. An invisible start node returns `not_found`.

Parameters not published.

pin_memoryPin a memory so it's never forgotten (writer+ for space memories)

Parameters not published.

promote_to_orgPromote a private memory to org visibility (not for space memories)

Parameters not published.

promote_to_spaceCopy one of your personal memories into a space (writer+). Optional `keep_original`.

Parameters not published.

set_core_preferencesSet user preferences for core memory categories

Parameters not published.

space_createCreate a Mind Space; you become its owner. `name`, optional `slug`, `org_id`, `description`, `forgetting_policy`, `llm_inference_enabled`, `show_service_names`.

Parameters not published.

space_erasePermanently delete a space and everything in it (owner, `confirm_slug`, optional `dry_run`)

Parameters not published.

space_getOne space, with member and memory counts (reader+)

Parameters not published.

space_listList your spaces with your role and counts. Optional `include_archived`.

Parameters not published.

space_member_addAdd a user or service with a role, or change an existing member's role (owner)

Parameters not published.

space_member_listList members (reader+)

Parameters not published.

space_member_removeRemove a member (owner; the last user owner can't be removed)

Parameters not published.

space_updateChange name, description, `forgetting_policy`, `llm_inference_enabled`, `show_service_names` (owner)

Parameters not published.

unpin_memoryUnpin a previously pinned memory (writer+ for space memories)

Parameters not published.

Ask AI

✦

Ask AI about Papez

Powered by Claude

Ask about Papez — answers draw on its README.

Security checks

6 checks have not been run yet.

Questions

About Papez

How do I install Papez?

Run uvx papez, then add the server to your MCP client's configuration.

Is Papez safe to use with an AI agent?

Its trust score is 58 out of 100 (fair). Conduid hasn't run static security checks on this repository yet, so review the source yourself before granting it credentials. It has no ConduID identity yet, so agent calls to it are not receipted.

Is Papez still maintained?

Yes — the last commit was 4 days ago. The repository has 31 stars and 5 forks.