Memora — Watchlist | MCP100 Index

MCP100 Assessment

Watchlist / Tracking

Why we’re tracking it

Follow memory storage and client integration changes as the product develops

Reviewed Oct 5, 2026 · Documentation review; complete setup workflow not independently tested

Best fit

Persistent collective memory for agents with graph relations and semantic retrieval

What we know

MCP memory layer with structured storage, semantic retrieval, graph relations, deduplication/supersession and optional cloud-backed storage.

Maintenance / activity evidence

Observed Sep 23, 2026 (UTC)release v0.4.6: contract kept byte-identical to memora-api-v1.0.0; bo…

Tracking evidence observed Oct 5, 2026. These dated signals do not establish current maintenance. Reviewed enrichment confidence: high.

Activity source · commits ↗Activity source · releases ↗

Verification gaps / watch-outs

Installation and comprehensive security have not been independently audited.
Memory-store scope and optional multi-database routing. Writes: Absorb/store/update memory; Cloud sync depending on backend.
Backend/provider-specific.
Sensitive access: Long-lived agent memory, Embedding/provider credentials

Default local SQLite is bounded; cloud sync and hosted embeddings materially expand exposure.

Security source · repository notice ↗

Full authority, access and evidence limits are available in Review context and Access & risk.

Popularity

725 repository stars

Stars observed Sep 25, 2026.

Preserve any already-reviewed adoption metrics in MCP100. Batch 4 did not refresh package downloads because the package-stats provider had insufficient credits.

All limitations / Review context

Which memory tools and data-retention boundaries are documented now?

The bounded assessment reviewed the official README, repository notices, security policy path, and security page for known material public-tracking blockers. No such blocker was identified. This is a public-tracking assessment, not an installation safety certification.

Default local SQLite is bounded; cloud sync and hosted embeddings materially expand exposure.
Memory-store scope and optional multi-database routing.
Backend/provider-specific.
SQLite by default with optional S3/R2/D1 and embedding providers.
Reviewed write actions: Absorb/store/update memory; Cloud sync depending on backend
Sensitive access: Long-lived agent memory, Embedding/provider credentials
Installation, client compatibility and comprehensive security have not been independently audited. Authentication and access depend on the configured services.
A project-specific SECURITY.md policy was not available at the fetched official path; the review remains limited to published repository notices.
No installation, runtime, client matrix, or complete tool-contract audit was performed in this review.
The fetched official SECURITY.md path returned HTTP 404; it was not treated as positive evidence.
Cloud backends and embedding options change data-flow boundaries.
Long-lived memories can contain sensitive context.

Original activity observations

2026-09-23T15:27:05Z: release v0.4.6: contract kept byte-identical to memora-api-v1.0.0; bo…

Technical identity

github.com/agentic-box/memora

Exact implementation source: https://github.com/agentic-box/memora ↗
Reviewed resource type: MCP Server
Repository identity: agentic-box/memora
Package identity: pypi:memora-mcp
Service endpoint: http://127.0.0.1:8910/mcp/<name>

Evidence & Sources

Evidence basis: Recorded source evidence
Setup runtime verification: Not independently runtime-tested by MCP100 for this setup workflow.

Reviewed Oct 5, 2026 · Latest source observation Oct 6, 2026

GitHub repository ↗
agentic-box/memora
Recorded repository material and project history.

Setup · Watchlist

Install & Agents

Publisher instructions.

Source evidence observed Oct 12, 2026.

Additional setup details required — see the specific gaps below.

Install / connect
Publisher instructions
For local stdio, install `pip install memora-mcp` and configure the client to launch `memora-server`. For remotely hosted HTTP, use the published container/service setup instead. Bare `memora` on PyPI is unrelated.
pip install memora-mcp
Configure / authenticate
Publisher instructions
Choose exactly one Memora storage mode before configuring the client: default local SQLite, explicit named multi-database, or cloud D1/R2. For the simple local stdio path run memora-server --no-graph in the selected client; cloud proxy credentials and stable endpoint are a different setup. No universal cloud credential JSON can be safely copied.
Check connection
Suggested check
After selecting storage, verify the memora MCP entry/tool listing in the client. For the publisher installer, check project MCP configuration; test only non-destructive memory search and do not create/alter memories just to verify startup.
Remove / uninstall / revoke
Remove the `memora` MCP entry from the client settings, then stop any container/service hosting it if applicable. If installed as a dedicated Python package and no longer needed, `pip uninstall memora-mcp`. Preserve database files by default.

Official Documentation & Sources

What it does

MCP memory layer with structured storage, semantic retrieval, graph relations, deduplication/supersession and optional cloud-backed storage.

Capability: Files & Knowledge
Resource type: MCP Server
Best fit: Persistent collective memory for agents with graph relations and semantic retrieval

Representative uses

Store/retrieve durable memories
Use semantic search and graph relations
Run multi-database service
Use local stdio or container HTTP

Limitations

Cloud backends and embedding options change data-flow boundaries.
Long-lived memories can contain sensitive context.

Reviewed access

Access & risk

Review the exact access boundary before connecting this MCP to accounts, credentials, or sensitive data.

Authentication: Backend/provider-specific.
Permissions: Memory-store scope and optional multi-database routing.
Data handling: SQLite by default with optional S3/R2/D1 and embedding providers.
Cost: free
Risk and cautions: Default local SQLite is bounded; cloud sync and hosted embeddings materially expand exposure.

Access scope and cautions

Write actions: Absorb/store/update memory Cloud sync depending on backend
Sensitive access: Long-lived agent memory Embedding/provider credentials
Risk level: moderate

FAQ

Based on reviewed evidence as of 2026-10-06. Watchlist tracking is not a recommendation or a security certification.

Which tasks are documented for this implementation?

The reviewed tasks include: Store/retrieve durable memories; Use semantic search and graph relations; Run multi-database service; Use local stdio or container HTTP.

What setup and authentication are documented?

Recorded installation/entrypoint step: pip install memora-mcp. Recorded transport: stdio, streamable_http. Prerequisites: Python. These are the documented setup details, not a claim that installation or client compatibility was independently tested. Authentication: Local mode keyless; cloud/embedding credentials as configured. A demo or keyless description is not independently verified read-only authority. Check the documented tool permissions rather than inferring authority from that label.

Can it modify data or perform write operations?

Documented write actions include: Absorb/store/update memory; Cloud sync depending on backend. Recorded permission boundary: Memory-store scope and optional multi-database routing.

What data-handling boundary should I consider?

SQLite by default with optional S3/R2/D1 and embedding providers. Recorded sensitive-access context: Long-lived agent memory; Embedding/provider credentials.

What limitations and risks should I check before use?

Default local SQLite is bounded; cloud sync and hosted embeddings materially expand exposure. Cloud backends and embedding options change data-flow boundaries. Long-lived memories can contain sensitive context. A project-specific SECURITY.md policy was not available at the fetched official path; the review remains limited to published repository notices.

Is this an official MCP?

Agentic Box is listed as Community implementation in MCP100’s current evidence.