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

No annotations are provided, so the description carries the full behavioral burden. It goes well beyond the schema by disclosing the layer-2 claim graph exclusion, the cluster-with-representative-chunks output shape, the LLM-extracted keyConcept diversification mechanism, the temporal ordering bias toward recent papers, and the single-paper dominance protection via maxClustersPerPaper. This is strong behavioral transparency.

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?

The description is front-loaded with the core purpose and each section adds distinct value. It is slightly repetitive in contrasting 'ranked list' with 'conceptual clusters' twice, and the keyConcept sentence is a bit dense, but overall it is compact for the amount of behavioral and usage context it provides.

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 complex discovery tool with 9 parameters and no output schema, the description covers the important non-schema aspects: scope, output type, use cases, alternatives, temporal bias, and diversification behavior. An agent has enough context to decide when to invoke it and to set date/maxClustersPerPaper meaningfully without needing additional explanation.

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 already high at 78%, so the baseline is 3. The description adds genuine value beyond the schema by explaining how to use dateTo/dateFrom for temporal slicing and how maxClustersPerPaper affects cluster diversity and paper dominance. Other parameters like concept, clusterCount, categories, and vectorModel are adequately covered by schema descriptions, though the prose does not add extra nuance for them.

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 uses a specific verb ('Map'), names the resource ('the paper corpus'), and clearly contrasts itself with sibling tools: it searches papers and chunks, not the claim graph, and returns conceptual clusters rather than a ranked list. This distinguishes it from methodist_explore_topic and search without needing to inspect sibling schemas.

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?

The description explicitly says when to use the tool ('what approaches exist to X' queries, breadth over depth) and when not to (for published claims, use methodist_explore_topic). It also gives concrete temporal guidance: specify dateTo for historical exploration and dateFrom+dateTo to slice a specific era, which is actionable routing and usage advice.

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