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

Annotations already flag destructiveHint=true and readOnlyHint=false, but the description adds real substance the annotations cannot convey: a keep decision records the choice without restoring permission, and an update writes a separate reference without altering accepted jobs or published posts. That is exactly the blast-radius information an agent needs for a destructive tool. Minor gap: it never says what is irreversible or whether the operation can be undone.

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 sentences, no filler, and the purpose is front-loaded. However the opening sentence is a comma-spliced noun list that is harder to parse than it needs to be, and the second sentence packs two distinct behaviors into one clause chain, slowing comprehension of the most safety-critical content.

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 six-parameter, five-required, destructive, nested-object tool this is middling: the decision outcomes are covered and an output schema exists to carry return values, but the meaning of work_id, request_key and the time-window parameters is never addressed, and the relationship between read and review operations on the same work_id is left implicit.

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 50%, so the schema documents read_skills, operation_id, publication_proposal_id and expected_catalog_version. The description usefully expands the decision enum (keep vs update) and ties expected_revision and fingerprint to a prerequisite read, but leaves work_id, request_key, link_id and the calendar from/after/until parameters untouched. Net it adds meaning beyond the schema on the decision axis without compensating 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?

The description names the resource ('linked work, spending, automations, calendar and source reviews') and the two operations (read vs review), but the opening noun pile is dense jargon: 'deduplicated linked work' is not defined anywhere, and the scope is not clearly distinguishable from siblings such as native.read_work or native.read_identity_projects. The second sentence does better by pinning down what a 'review' actually does, which is the real differentiator.

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?

'Review a frozen source only for an explicit creator decision' gives a genuine gating condition for the mutating path, and 'Read the exact link revision and fingerprint first' establishes a sequencing prerequisite. What's missing is routing to alternatives — nothing tells the agent when to use native.read_work or native.identity_brand_source instead, despite many read-oriented siblings.

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