diagnose_mcdev_retrieve_logs

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

Goes far beyond the readOnly/openWorld/destructive annotations: it discloses exactly what is read (bounded project/logs timestamp files, never error companions or auth/session content), what is excluded (no subprocess, network, native execution, elicitation, rerun), and what is returned (finite observations and basename/time provenance, never raw logs, paths, errors, secrets). It also flags the --noLogFile no-log case and unknown-record suppression, which an agent could not infer from annotations.

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 scope are front-loaded in the first sentence, and every clause carries constraints or exclusions rather than filler. However, several sentences are densely semicolon-packed run-ons (the window/format sentence crowds five restrictions together), making it harder to scan than it needs to be.

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 no-output-schema, seven-required-parameter diagnostic tool, it covers the essentials: quiescence, window bound, selector syntax, negative-evidence semantics, and the shape of the result ('finite observations and validated basename/time provenance'). The main residual gap is that the structure of those 'observations' is never characterized, so the agent must call blind on return shape.

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 0% schema description coverage, the description has to carry all seven parameters and it mostly does: it pins credential/businessUnit to 'one exact configured' value, constrains metadataType to canonical retrieve query/script keys, and defines the startedAfter/startedBefore pair as a UTC ISO millisecond window of at most 24 hours. It also warns against options, aliases or quotes in selectors. workspaceRoot is never mentioned and exact key format is only implied, so it is not fully compensatory.

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: 'Read-only historical normal-log diagnosis for one exact configured credential/businessUnit and query or script key.' That is precise enough to distinguish it from run_mcdev_command and reconcile_mcdev_operation, which operate on live operations rather than historical log files. It stops short of explicitly naming which sibling to prefer, keeping it at 4.

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 real usage constraints: 'Requires caller-managed quiescence' and 'absent evidence is no-match, not a reason to repeat retrieval', plus 'investigate manually when evidence is insufficient'. These tell the agent when this tool applies and when to stop, but no sibling tool is named as an alternative for those cases.

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