ges

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

The annotations only declare readOnlyHint=true and openWorldHint=false. The description goes far beyond that by enumerating statistical assumptions (Causal Markov, faithfulness, causal sufficiency, acyclicity), pre-conditions, failure modes with likely causes and remedies, and a typical minimum sample size. This is precisely the kind of behavioral disclosure that helps an agent anticipate edge cases. No contradiction with 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?

The description is a single dense paragraph covering validation, assumptions, pre-conditions, failure modes, alternatives, and minimum N. Every sentence carries useful information, but the lack of bullet points or section breaks makes it harder to scan quickly. It is not excessive in length, but structure could be improved.

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 causal discovery tool, the description covers validation tier, assumptions, data requirements, failure modes with remedies, and alternative tools. The output schema exists so return values need not be explained. It lacks an example or a more explicit statement of what a typical result looks like, but the provided context is quite complete for an agent to decide whether to call this tool and how to react to common failure modes.

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 every parameter already has a detailed description in the schema. The tool description does not add parameter-specific semantics beyond the schema; it only mentions sample size requirements in a general way and does not discuss data_path, max_iter, detail, etc. Baseline 3 is appropriate.

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 opens with 'Greedy Equivalence Search', which essentially restates the tool name, and then dives into assumptions and failure modes without ever explicitly stating that this tool performs score-based causal discovery on observational data. An agent familiar with GES may infer the purpose, but there is no direct 'perform GES to estimate a causal DAG' statement, and the differentiation from siblings like pc_algorithm or fci rests on the algorithm name rather than an explicit comparison.

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 provides clear pre-conditions (i.i.d. observational data with enough samples for CI tests; labelled data for invariance-based methods), a typical minimum N, and a concrete conditional alternative: switch to FCI if latent confounders are plausible. It lists alternate tools, but does not explicitly say when to choose GES over PC or LiNGAM. Overall it gives solid context and at least one explicit exclusion, so not quite a full 5.

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