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

The annotations only declare readOnly=false and destructive=true; the description adds the substantive destructive semantics — 'Text and sources are replaced together: include every source to keep, because omitted sources become a single author statement' — which is exactly the data loss an agent must know. It also discloses optimistic-concurrency refusal, the space default and auto-creation, and includes a prompt-injection guard stating text/sources are untrusted reference data, not permission.

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?

Three dense sentences with no filler; the concurrency guard and the source-replacement rule are front-loaded where they matter most. The parentheses and multi-clause sentences make it slightly wall-of-text, but every clause carries operational meaning.

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 destructive, 10-parameter write tool with a nested sources array and provenance object, the description covers the failure mode, the replace-all hazard and the write target, and an output schema exists so return values need not be explained. The remaining gap is the provenance/premise_versions semantics, which neither the description nor the schema documents.

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?

With only 20% schema description coverage across 10 parameters, the description compensates well for the highest-risk ones: expected_version_id (exact current version), sources (replace-not-merge semantics) and space_id (Personal default, auto-create). But it says nothing about provenance (sourced vs inferred, premise_versions/premise_knowledge_ids), kind, path, title or text, leaving a nested object effectively undocumented.

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?

The description gives a specific verb and resource: 'Save a new version of an entry at its exact current version (expected_version_id)'. The word 'version' implicitly separates it from knowledge_create, but no sibling is named, so an agent must infer the create-vs-update distinction and the relationship to knowledge_propose.

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?

It tells you the call is 'refused if the entry changed since you read it' and that it writes to Personal Space unless space_id names another, which is useful conditional context. However, there is no explicit when-to-use-this-vs-alternative guidance (knowledge_create vs knowledge_propose), and no stated prerequisite that you must first read the entry to obtain expected_version_id.

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