notears

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

Annotations already mark the tool read-only and not open-world, and the description adds non-obvious behavioral context: failure modes (unstable skeleton / undirected edges), a likely diagnosis (faithfulness or sample size), remediation steps, and a minimum sample-size heuristic. This goes beyond the structured annotations without contradicting them.

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 lead sentence is sharp and the labels (Validation, Assumptions, Pre-conditions, Failure modes, Alternatives, Typical minimum N) make it scannable, but parts of the body is boilerplate from other causal-discovery methods (PC/GES/FCI/LiNGAM) and the 'validated evidence tier' sentence is vague. These weaken the otherwise compact structure.

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 13-parameter tool with a full output schema, the description covers the main non-schema context an agent needs: sample-size expectations, failure modes, data preconditions at a high level, and which alternative tool to switch to. It is not complete regarding NOTEARS-specific statistical assumptions, but schema and output schema carry the rest.

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% and each of the 13 parameters has a meaningful schema description (e.g., w_threshold pruning, as_handle caching, detail payload depth). The tool description itself adds no parameter-level meaning, so a 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 opening sentence is a specific verb+resource+algorithm statement: 'Learn a DAG from data using NOTEARS (Zheng et al. 2018).' This distinguishes it from DAG-learning siblings like pc_algorithm, ges, and fci by naming the exact method, even though later assumptions text is generic.

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 lists Alternatives and includes one conditional routing rule ('switch to FCI if latent confounders are plausible'), plus a typical minimum N. However, it never states the conditions under which NOTEARS is preferred over pc_algorithm, ges, or lingam, and the pre-conditions/assumptions section reads like a generic causal-discovery template rather than NOTEARS-specific guidance.

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