remember

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

Goes well past the annotations (readOnlyHint=false, destructiveHint=false, idempotentHint=false): it discloses the four per-item outcomes including near-duplicate merging and topic conflicts with a required user consultation, and explains that supersedes_id causes the prior entry to become superseded rather than deleted. That is exactly the 'what actually happens to my data' context an agent needs before writing.

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?

Front-loads the core action and keeps everything in one tight block with no filler. The long comma-separated enumeration of knowledge kinds is dense but each item earns its place by disambiguating what to save.

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?

For a write tool with four params, an output schema, and partial annotation coverage, the description covers timing, batching limits, return statuses, conflict handling, and the replacement/confirmation workflow. An agent could invoke it correctly and interpret the result without opening the schema.

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 75% and already documents most fields, yet the description still adds value: it explains `supersedes_id` as the replacement mechanism and `confirm` as the resolution path for ai_suggested ids. It does not explain the interplay between `items` and `confirm` in a single call, or the required `project` slug beyond the schema's own note.

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?

Opens with a specific verb+resource ('Save project knowledge') and immediately enumerates exactly what counts as knowledge (decisions, rejected approaches, constraints, bugs, fixes, risks, open questions). The batch contract ('up to 50 items') and the per-item outcome vocabulary make it unmistakable next to siblings like search_memory or checkpoint.

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?

Gives explicit timing guidance ('AS YOU GO (not only at the end)') and routes two cases to siblings: conflicts go to the user then resolve_conflict, and ai_suggested items get confirmed via the `confirm` param once the user agrees. It does not state when NOT to use this (e.g. versus checkpoint or import_context), which keeps it short of a 5.

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