AgentDB — Watchlist | MCP100 Index

MCP100 Assessment

Watchlist / Tracking

Why we’re tracking it

Follow-up value is specific to AgentDB vector and episodic memory tools available over MCP: Do memory contracts and storage formats remain interoperable during cognitive-container and MCP tool changes?

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

Best fit

Local vector/graph memory and self-learning retrieval for agents.

What we know

Single-file vector memory SDK/CLI with semantic and hybrid search, learning state and an MCP server exposing memory/pattern/causal/skill operations.

Maintenance / activity evidence

agentdb 3.0.0-alpha.20 published 2026-07-30T01:00:04.258Z

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

Verification gaps / watch-outs

Installation and comprehensive security have not been independently audited.
Local configured DB/file scope. Writes: Store/update/delete memory; Import/export.
Authentication not yet verified.
Sensitive access: Agent memory, Embeddings/metadata

Local storage reduces exposure, but the MCP toolset can mutate and export persistent memory.

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

Popularity

· 335k package downloads / 30 days

Downloads observed Sep 25, 2026.

Do not overwrite existing reviewed stars/download metrics with nulls from this enrichment batch.

All limitations / Review context

Do memory contracts and storage formats remain interoperable during cognitive-container and MCP tool changes?

The bounded review of the publisher's documented AgentDB vector and episodic memory tools available over MCP and release notices identified no disclosed material condition preventing an accurately scoped public tracking entry. The resource's operational permissions remain uncertainty; this assessment does not certify installation safety.

Local storage reduces exposure, but the MCP toolset can mutate and export persistent memory.
Local configured DB/file scope.
Data can remain in a local .rvf container.
Reviewed write actions: Store/update/delete memory; Import/export
Sensitive access: Agent memory, Embeddings/metadata
Installation, client compatibility and comprehensive security have not been independently audited. Authentication and access depend on the configured services.
Runtime behavior, dependencies and undisclosed vulnerabilities were not independently audited.
Documented tool authority and any remote data handling must be checked before installation or use.
Tool inventory, client matrix and permissions completeness were not requirements for this Radar tracking review.
The package combines SDK, CLI and MCP roles.
Security feature claims vary by version and need config-specific interpretation.

Original activity observations

agentdb 3.0.0-alpha.20 published 2026-07-30T01:00:04.258Z

Technical identity

npm/agentdb

Exact implementation source: https://www.npmjs.com/package/agentdb ↗
Reviewed resource type: SDK
Repository identity: ruvnet/agentdb
Package identity: agentdb

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 7, 2026

Package registry ↗
agentdb
Recorded package identity, distribution or download evidence.
Includes references whose attribution is not established.
GitHub repository ↗
ruvnet/agentdb
Recorded repository material and project history.
Includes references whose attribution is not established.

Setup · Watchlist

SDK setup

Publisher instructions.

Source evidence observed Oct 14, 2026.

Additional setup details required — see the specific gaps below.

Warning

Not runtime tested, no third-party installation, network authorization, paid calls, or mutation executed.
Do not claim production service is available based on README alone.
SDK also offers explicit MCP mode; npm version is alpha; local storage/permissions may be required
MCP start subcommand documented, but npm page is alpha; pin a tested version if stability matters.
Official npm publisher package page README inspected for this exact npm identity; static research only. Some setup statements are conservatively summarized; no source code or live client execution checked. V2 review did not run any live client.
Check actual host syntax and supply credentials in secure client environment; do not paste placeholders verbatim.
Avoid write, payment, token transfer, publishing, sending, booking and remote account changes as tests.
Replace credential, action and registration placeholders; follow the matching publisher/client prerequisites. No execution performed.
Use the matching host, actual registration name and original installation environment; this does not revoke credentials.

Install SDK / application integration
Publisher instructions
SDK also offers explicit MCP mode; npm version is alpha; local storage/permissions may be required | Publisher npm README revisited on 2026-10-14; exact identity retained.
Reviewed setup example · Claude Code only · Publisher instructions
claude mcp add agentdb -- npx agentdb@latest mcp start
Configure / authenticate
MCP100 setup guide
Use `npx agentdb@latest mcp start` for MCP; `npm install agentdb` is a separate library path.
Check connection
Suggested check
Check for agentdb_* tool names in the client without writing to a real database.
Remove / uninstall / revoke
MCP100 setup guide
If registered by the publisher Claude Code example under `agentdb`, run the provided unregister command; for a different MCP client or name, delete the actual client registration. Removing a local client entry is distinct from deleting a vendor account/token.
Remove named client registration · Claude Code only · MCP100 composed guide
claude mcp remove agentdb

Official Documentation & Sources

What it does

Single-file vector memory SDK/CLI with semantic and hybrid search, learning state and an MCP server exposing memory/pattern/causal/skill operations.

Capability: Data & Databases
Resource type: SDK
Best fit: Local vector/graph memory and self-learning retrieval for agents.

Representative uses

Embed vector memory in applications
Search/store patterns and memories
Run MCP memory tools
Import/export and maintain self-learning state

Limitations

The package combines SDK, CLI and MCP roles.
Security feature claims vary by version and need config-specific interpretation.

Reviewed access

Access & risk

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

Authentication: Not documented in the reviewed sources
Permissions: Local configured DB/file scope.
Data handling: Data can remain in a local .rvf container.
Cost: free
Risk and cautions: Local storage reduces exposure, but the MCP toolset can mutate and export persistent memory.

Access scope and cautions

Write actions: Store/update/delete memory Import/export
Sensitive access: Agent memory Embeddings/metadata
Risk level: moderate

FAQ

Based on reviewed evidence as of 2026-10-07. Watchlist tracking is not a recommendation or a security certification.

Which tasks are documented for this implementation?

The reviewed tasks include: Embed vector memory in applications; Search/store patterns and memories; Run MCP memory tools; Import/export and maintain self-learning state.

What setup information is documented?

Recorded installation/entrypoint step: npm install agentdb; claude mcp add agentdb -- npx agentdb@latest mcp start. Recorded transport: stdio. Prerequisites: Node/npm. These are the documented setup details, not a claim that installation or client compatibility was independently tested.

Can it modify data or perform write operations?

Documented write actions include: Store/update/delete memory; Import/export. Recorded permission boundary: Local configured DB/file scope.

What data-handling boundary should I consider?

Data can remain in a local .rvf container. Recorded sensitive-access context: Agent memory; Embeddings/metadata.

What limitations and risks should I check before use?

Local storage reduces exposure, but the MCP toolset can mutate and export persistent memory. The package combines SDK, CLI and MCP roles. Security feature claims vary by version and need config-specific interpretation. Runtime behavior, dependencies and undisclosed vulnerabilities were not independently audited. Documented tool authority and any remote data handling must be checked before installation or use.

Is this an official MCP?

RuvNet is listed as First-party MCP in MCP100’s current evidence.