Skip to content
Agentmart
Submit a tool

Memharness

Bi-temporal, provenance-carrying long-term memory for AI agents, in one SQLite file.

Connector by las7 · GitHub account verified by the MCP Registry

IndexedApache-2.0Local (stdio)★ 6
Write a review

Install

Works with Claude Code, Codex, VS Code, Cursor, Windsurf, Claude Desktop. Read the repository before you give any tool access to your data.

npx agentmarket install connector/las7-memharness

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Bi-temporal, provenance-carrying long-term memory for AI agents, in one SQLite file.

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this server can do and what it asks for, from its tool list and install config. Checked Oct 25, 2026.

Tools
3 read · 2 write · 0 delete · 2 not yet classified
Secrets
None asked for
Install scripts
Not scanned yet
Runs as
Node.js package

Tools by what they do 7

Read from the source code; the running server may list different tools.

Read 3 · Look things up without changing anything.

  • diff
  • recall
  • stats

Write 2 · Create, change, send or run things.

  • remember
  • revise

Not yet classified 2 · We could not tell from the name and description.

  • forget
  • why

Secrets and settings it asks for

Other settings (environment variables or headers).

  • MEMHARNESS_DB

Tools (7)

Read from the source code; the running server may differ. Descriptions are the server’s own words; hidden characters are shown as [U+…].

  • diff

    What changed in memory since a given ISO date: new facts learned, beliefs revised, facts retracted. The 'what have you learned about X since Monday?' query.

  • forget

    Retract a fact by id, or retract EVERYTHING from a given source_ref (provenance-based deletion: 'forget everything from that session'). Tombstoned, not erased: as_of queries before the retraction still show history; current recall never returns it.

  • recall

    Retrieve relevant memories. A plain call returns current beliefs ranked by relevance × confidence × importance × recency (recency is freshened each time a fact is recalled). as_of (ISO date, e.g. '2026-06-01') returns what was believed AT THAT TIME — including facts since revised or retracted. Use as_of to answer 'what did you think before I corrected you?'. subject filters to one entity; kind filters to semantic/episodic/procedural.

  • remember

    Store ONE atomic fact in long-term memory: a single assertion, ideally one sentence (aim for under 250 characters; writes over 1500 are rejected). Split compound knowledge into multiple remember calls so each piece can be revised independently later. Use for durable knowledge about the user, their projects, preferences, decisions, and environment — not transient task state. If this contradicts an existing belief, find it with recall and use revise instead. Always fill source_ref with where this came from (session, file, URL) if known. If the fact describes code you just read at a known commit,…

  • revise

    Replace a belief with an updated one. The old fact is closed and linked to its successor — history is preserved, never deleted. Use when the user corrects you or circumstances change. Find the old fact's id with recall first. valid_from can backdate when the new state actually began (e.g. 'moved last month'). importance and kind are inherited from the old fact unless you override them.

  • stats

    Memory database statistics: counts, top subjects, schema version, db path.

  • why

    Full provenance and revision chain for a fact: where it came from, when it was learned, what it superseded, and what superseded it. Use to answer 'why do you believe that?' or to audit a belief before revising it.

Signals

Stars
6
Forks
1
Open issues
0
Weekly downloads
14
Last commit
Jun 26, 2026 (4 months ago)
Last release
Jun 20, 2026 (4 months ago)
Last package publish
Jun 20, 2026 (4 months ago)

From GitHub (fetched Oct 24, 2026) and the package registry.

Details

Registry name
io.github.las7/memharness
Version
0.1.2
Auth
No auth
Packages
npm @memharness/mcp
License
Apache-2.0
First seen
Oct 24, 2026
Last checked
Oct 24, 2026
Cite this page

Memharness MCP server. Agent Mart. https://agentmart.dev/connectors/las7-memharness (last updated 2026-10-24)