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

Annotations already declare readOnly/no-destructive, so the safety bar is covered. Beyond that the description adds real behavior: newest-first ordering, that a score only appears once the run is DONE, and an anti-polling operating constraint. It does not disclose result limits or pagination, which 'Recent runs' hints at but never defines.

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?

Purpose and returned fields are front-loaded, with behavioral guidance following. The capitalized 'THE CHECK-BACK' phrase for emphasis is slightly informal, but every sentence carries usable information and none is padding.

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 list tool with full annotation coverage but no output schema, the description compensates by describing the returned fields (status, score, label, timestamps) and their lifecycle. An agent has everything needed to call it and interpret the result correctly.

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 description coverage is 100%, so both evalId and workspace are already documented in the schema. The description's 'for one eval' aligns with evalId but adds no format, scope, or defaulting detail beyond what the schema provides. Baseline 3 is correct.

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?

States a specific verb and resource ('Recent runs for one eval, newest first') and enumerates the returned fields including the full status lifecycle (PENDING/RUNNING/SCORING/DONE/FAILED). It clearly distinguishes itself from the run-triggering tool by calling itself 'THE CHECK-BACK for caliper_evals_run', though it leaves any contrast with the sibling caliper_evals_runs_get (single-run fetch) implicit.

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 names the selection condition ('when the user asks how the run went, read this') and ties the tool to a specific sibling. It also gives a negative guideline ('don't poll in a loop; check when the user asks or when reporting'), which is exactly the when-not-to-use guidance an agent needs.

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