Skip to main content
Glama
worakorn-prince

th-memory-mcp

by worakorn-prince
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Local

Extract memories from interactions

extract_memories

Scans recent captured interactions for memory-intent phrases and proposes candidate memories, with a dry-run default and optional apply mode to create them.

Instructions

Scan recent captured interactions for memory-intent phrases and propose memory candidates (deterministic, no LLM). Dry-run by default; pass apply=true to create them (source=captured).

Input Schema

NameRequiredDescriptionDefault
kindNoInteraction kind to scan (default prompt)
applyNoCreate the proposed memories (default false = propose only)
limitNoMax recent interactions to scan (default 50)
userIdNoScope extracted memories to a user (USER scope)
sessionIdNoFilter by captured session id (default: all sessions)

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changedv2.4.0
    • addedInput schema / properties / sessionId
      Added value: +{
+  "description": "Filter by captured session id (default: all sessions)",
+  "maxLength": 200,
+  "type": "string"
+}
  2. First observedv2.2.1

TDQS

A4.2/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full burden and mostly does: it discloses the non-mutating default, the mutation trigger (apply=true), the mechanism (deterministic, no LLM), and the provenance stamped on created records (source=captured). It omits deduplication behavior on repeated runs and what the proposal output contains.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Two sentences, zero filler, and the safety-critical fact (dry-run by default) arrives before the opt-in mutation flag.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Five parameters, no output schema, and no annotations, yet the description still conveys the scan target, safety default, and write side effect adequately. It could be slightly more complete about what a 'memory candidate' looks like and whether re-applying creates duplicates.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so kind, apply, limit, userId, and sessionId are already documented in the schema. The description adds only the source=captured provenance tied to apply, which is a genuine but small increment beyond structured data; baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource ('scan recent captured interactions for memory-intent phrases and propose memory candidates') plus the mechanism ('deterministic, no LLM'). The propose-vs-create distinction implicitly separates it from sibling tools like remember and get_recent_interactions.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives clear operating context: default is dry-run, and apply=true is the condition for persistence. It does not name alternatives or exclusions (e.g., when to use remember instead, or how this relates to get_recent_interactions), so the agent must infer the boundary.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Deploy Server

Other Tools