Add a memory

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

Annotations already mark this as destructive/non-idempotent, but the description adds far more: secret scrubbing, a Jev gate that can refuse with reasons, kind-correction behavior, and the dead-end dedup/update rule. These are non-obvious behavioral traits an agent could not infer from structured fields.

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

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

The core action is front-loaded in the first sentence, followed by the gating rules, then the critical 'do not call' guidance. It is dense and the dead-end paragraph is long, but each sentence states a concrete rule with behavioral consequence rather than filler.

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

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

Given a write tool with no output schema and a validation gate, the description covers refusal outcomes, kind correction, and the auto-record interaction well. What a successful append returns is only implied ('append one memory line'), a minor gap.

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

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

Schema coverage is 0%, so the description carries the load. It adds real meaning for the 'dead-end' kind (must say what was tried and why, dedup semantics) but leaves the other six enum values and the text constraints (maxLength 500, minLength 3) unexplained. Partial compensation for the coverage gap.

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

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

Starts with a specific verb+resource and scope: 'Append one memory line to JEVMEM.md.' It is clearly distinguishable from the read-only siblings (list_memory, search_memory, audit_memory) because it describes a single-line write with validation, not a query.

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

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

Explicitly states when NOT to use it ('do not call this to repeat what the user just said; call it only when the user asks you to record something the conversation itself does not state') and explains the auto-recording context that determines that. This is a genuine when/when-not routing rule an agent can act on.

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