Field-by-field comparison

Compare the evidence before choosing an MCP.

Each signal stays separate, missing facts remain visible, and the column order follows your selection.

MCPs in this comparison

Uncheck an MCP to remove it. Comparing two or three keeps each signal in its own row.

Open canonical comparison link

Comparable alternatives

Shown only when capability or industry context overlaps.

Side-by-side MCP evidence comparison
Evidence fieldCogneeShortlist candidateMCPVault for ObsidianMCP100 Recommended
Task fit
Best for
  • Teams that want persistent graph-based memory/knowledge retrieval across AI sessions and can govern stored data.
  • Obsidian users who want local search and supervised note maintenance
  • Vaults protected by version control or reliable backups
  • Users able to expose only a dedicated, bounded vault
Not ideal for
  • Workloads that require stateless operation or cannot persist project/user information in a memory graph.
  • Unbacked-up personal knowledge bases
  • Read-only research where a simpler filesystem search is sufficient
  • Enterprise knowledge requiring centralized authorization and audit controls
Avoid when
  • Do not ingest sensitive data without understanding where the selected Cognee deployment stores graph/vector/content data and who can query it.
  • The vault contains secrets, regulated data, or material that cannot reach the selected AI provider
  • Write, move, and delete tools cannot be confirmed by a human
  • The vault path cannot be narrowed to a dedicated directory
Provenance
Provenance details

First-party MCP

github:topoteretes · Publisher source ↗

Community implementation

BitBonsai · Publisher source ↗

Maintenance
Maintenance details

Repo: Not documented

Package: Not documented

Repo: Sep 21, 2026

Package: Not documented

Popularity evidence
GitHub starsNot documented

1,672

GitHub stars · checked 2026-09-24T04:14:39.538000Z

Package downloads / 30 days

1,288

package downloads / 30 days · checked 2026-10-21

No reliable download data
Shared repository stars

32,633

Shared repository stars (not MCP-specific) · checked 2026-10-21

Not documented
Client coverage
Client coverage detailsNot documented
Claude Code✓ Local stdio
Claude Desktop✓ Local stdio
Codex✓ Local stdio
VS Code✓ Local stdio
Cursor✓ Local stdio
OpenCode✓ Local stdio
Client coverage noteLocal stdio runs on your computer. Streamable HTTP connects to a remotely hosted MCP server.
Access and data
AuthenticationCloud-backed configuration can use Cognee API key headers; self-hosted HTTP deployments can use documented Bearer-token/auth modes.None at the local stdio layer.
CostOpen-source/self-hosted infrastructure or Cognee cloud plan plus model/provider costs may apply.MCPVault is MIT licensed and free to run locally. MCP-host, model, backup, and storage costs remain.
PermissionsCan ingest, persist, query and in some modes delete/manage memory datasets according to deployment credentials.The server has filesystem read and write access within the supplied vault root, including note creation, overwrite or patch, frontmatter changes, moves, tags, and deletion. Operating-system permissions and the configured vault path are the primary boundary.
Data handlingContent supplied to Cognee can be persisted in graph/vector/storage backends; retention and tenant isolation depend on the deployment.Vault content and metadata are read locally and returned to the MCP client and model. Changes are written directly to local files. MCPVault states that it does not receive, store, or transmit vault data, while the selected AI provider still processes content sent by the client.
Limitations
Tradeoffs
  • Promoted after strict re-audit: canonical identity and MCP-native/publisher-backed implementation are sufficiently evidenced, with real workflow value, usable documentation, understandable permissions, plausible maintenance, and no unresolved security veto. Still not strong enough for Recommended.
  • Local stdio avoids a hosted MCP service, but note contents still enter the MCP host and model context.
  • Frontmatter-aware writes are more useful than generic file access but increase integrity risk.
  • npx makes setup simple, while stricter environments may prefer a pinned and reviewed package version.
  • Version 0.16.0 migrated to MCP TypeScript SDK v2 / protocol 2026-07-28 while retaining dual-era stdio compatibility.
Risk contextPersistent memory can retain sensitive project/user information beyond the current conversation.A selected vault can contain a large body of private knowledge, and write tools can overwrite, move, or delete it. Use a dedicated vault, reliable backups or version control, an absolute path, a reviewed package version, and explicit approval for writes.
Evidence date
Editorial review2026-10-212026-09-05
Candidate evidence2026-10-21T00:00:00Z2026-09-24T04:25:05.554391Z

Popularity, maintenance, fit, permissions, and client support are independent evidence fields.

Read the evidence method →