Skip to main content
Glama

memory_propose

Idempotent

Propose a durable Lucy memory as a pending write for human review, supporting create or replace operations. Success means pending, not saved, so nothing goes live without approval.

Instructions

Propose a durable Lucy memory write. Success means pending, not saved. Do not send secrets. Humans approve on Lucy unless they enabled auto memory.

Input Schema

NameRequiredDescriptionDefault

No arguments

Output Schema

NameRequiredDescriptionDefault
idYes
resultYes
statusYes
operationYes
relationsYes
requestedYes
created_atYes
decided_atYes
expires_atYes
schema_versionYes
proposal_versionYes

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv0.1.0

TDQS

A3.8/5.0
Behavior4/5

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

The annotations already state that the tool is not read-only, is idempotent, is not destructive, and is open-world. The description adds information the annotations do not carry: "Success means pending, not saved," which tells the agent that a pending state follows the call and that later human approval may be needed. It also warns against secrets and notes the auto-memory exception. It does not describe what happens to a proposal when it is rejected or when it expires, or how replace operations affect version conflicts.

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?

The description is three short sentences, and each one carries weight: purpose, the pending-vs-saved outcome, the secrets prohibition, and the approval model. The critical state caveat is placed immediately after the purpose statement, so the reader hits it first.

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?

The output schema exists, so the description does not need to explain return values. For a tool with oneOf create and replace branches, idempotency, and a human-approval gate, the description covers the workflow shape, the safety constraint, and the approval requirement well. It leaves out the replace and versioning semantics, which is a real gap given the complexity, but the remaining gaps are secondary.

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 description coverage is reported at 100%, so the baseline is 3. The description adds no meaning about the actual payload parameters: the key, kind, scope, content, justification, idempotency_key, and the replace variant's memory_id and expected_version are all left to the schema. It does not even hint that there are two shapes: create and replace with optimistic concurrency.

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

Purpose4/5

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

The description names a specific verb and resource: "Propose a durable Lucy memory write." That verb clearly separates it from sibling tools such as memory_recall, which reads memory, and memory_proposal_status, which checks proposal state. It never explicitly names or contrasts those siblings, so the differentiation stays implicit rather than stated.

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

Usage Guidelines3/5

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

"Propose a durable Lucy memory write" and "Do not send secrets" imply when this tool should be used: you record durable memory and you keep sensitive data out. The description also says humans approve the proposal, so the agent knows a second step follows. It never says when to prefer this tool over memory_recall or memory_proposal_status, and it gives no exclusion rules beyond the secrets warning.

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