Memharness MCP server

memharness

<!-- mcp-name: io.github.las7/memharness -->

A bi-temporal, provenance-carrying memory primitive for AI agents. One SQLite file. No LLM or network calls in the storage layer. Exposed to any agent via MCP.

Most agent memory is a bag of strings. memharness stores facts, and combines three semantics that incumbents tend to split apart:

(validfrom/validto) separately from when the agent learned it (txat). So you can ask: "what did you believe on March 1st?"*

1. Bi-temporal: every fact records when it became true in the world

to its successor. "What did you think before I corrected you?" has an answer.

1. Supersession, never deletion: corrections close the old fact and link it

"Why do you believe that?" has an answer. So does "forget everything from that session."

1. Provenance per fact: every memory cites who said it, where, and when.

The storage layer is deterministic: no LLM, no network, no background daemon. It's plain SQLite, so you can open the file with any client.

Run it yourself: cd examples &amp;&amp; npm install &amp;&amp; npm run demo

When to use this (and when not to)

memharness is not a magic accuracy upgrade, and it is honest about that. If your agent's memory is small and static and comfortably fits the context window, a CLAUDE.md file (or just stuffing the history into the prompt) is simpler, and on short histories full context will match or beat any external memory system.

Reach for memharness when:

want to (or can) paste into every prompt.

• History outgrows the window: months of facts, many subjects, more than you

decision?" (asof), "what changed since Monday?" (diff), "why does it believe this?" (why). These are queries a bag of strings cannot answer.

• You need an audit trail: *"what did the agent believe when it made this

session/file/source"* in one call (GDPR-shaped, not a string search).

• You need provenance-scoped deletion: *"forget everything from that

overwrite, so old reasoning stays explainable.

• Beliefs change over time: corrections should supersede, not silently

How it compares

Honest, and pointed at the thing memharness actually does differently: it is a deterministic, auditable storage layer rather than an extraction service.

Where the others win, plainly: mem0 and Zep do automatic fact extraction from raw conversation, which memharness deliberately does not (the write path stays model-free; a client or skill decides what is worth remembering). Plain CLAUDE.md needs no install at all. memharness earns its place when you need the temporal and provenance queries the others don't offer.

Packages

Quick start (MCP)

Shortened. The full README is on GitHub.