Compare

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

With no annotations at all, the description must carry the full behavioral burden. It does disclose what the tool computes (chroma_sim, rhythm_corr, band dB deltas, log-mel L1, worst bars, advice) and hints at staleness via render's last-run reply, but it never states that this is a read-only analysis that does not alter the project, render, or reference. For a no-annotation tool, that omission is a real gap even though the name implies non-destructiveness.

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?

Dense but purposeful: purpose, output scores, stem syntax, and usage conditions are all front-loaded in roughly two sentences. The phrase 'render's reply says when it last ran' is slightly oblique and the 'plus the worst bars and advice' clause jams return detail in, but there is no filler and every clause contributes.

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?

An output schema exists, so return values need not be explained, yet the description still outlines them helpfully. The unresolved gaps are the undocumented bars, detail, and offset_b parameters and the lack of any explicit statement about side effects or permissions. For a 6-parameter, 0%-schema-coverage tool, that leaves an agent with meaningful uncertainty before calling it.

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 0%, so the description must compensate, and it does add genuine meaning for the two most important parameters: a and b default to 'render' and 'ref', and stem targeting is shown with a='track:bass', b='ref:bass'. However, three of six parameters — bars, detail (default 8), and offset_b (default 0) — receive no mention at all, leaving their semantics to guesswork.

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: 'Compare two sources bar by bar on the project grid', with the default comparison pair (render vs reference) spelled out. It does not name the cmp_* siblings, so an agent must infer that this is the general bar-level comparator rather than the summary/section/bars variants, but the core action is unambiguous.

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 triggers: use when a reference is set (as target or inspiration) and a draft is about to go to the person, or when feedback says it does not match the reference. It also notes that 'render's reply says when it last ran', which supports deciding whether to re-run. No explicit exclusions or named alternatives (cmp_summary, cmp_sections, cmp_worst) are given, so it stops short of full routing guidance.

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