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

Annotations only declare the safety profile (readOnly, idempotent, non-destructive); the description adds substantial behavioral context the annotations cannot: outcome precedence (Stop > Recycle > Hold > Conditional go > Go), the rule that fired, waiver handling, Hold-vs-Conditional-go triggers, modifier semantics, and the receipt-hash return. This is exactly the beyond-annotations disclosure the dimension rewards.

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?

Front-loaded with the core purpose followed by the outcome precedence, then edge cases. It is dense and long but nearly every sentence carries decision-relevant information (precedence, waiver rules, G8 exception). Slightly prose-heavy for a single paragraph but no filler.

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 12-parameter tool with nested context, no output schema, and high branching logic, the description covers the decision model, modifiers, and even mentions the return value ('receipt hash of the inputs'). Complete enough to call correctly; only the exact receipt contents and warning envelope details are left implicit.

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 92% schema coverage the baseline is 3, but the description adds real meaning for several parameters: business_case_valid=false → Stop, earlier_stage_invalid=true → Recycle with a named return stage, and work_started_beyond at G8 meaning Hold rather than a modifier. These are semantic rules not captured by the schema, lifting it above baseline.

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+resource+scope: 'Evaluates gate g.g0 … g.g10 against your evidence.' An agent knows exactly what the tool consumes and produces. It does not, however, explicitly differentiate itself from siblings like validate_plan or check_mix, so it lands at 4 rather than 5.

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?

Provides concrete usage guidance: 'Call without evidence to list what the gate needs,' and tells when to pass earlier_stage_invalid, work_started_beyond, and invalid_stage. It never states when to prefer a sibling tool (check_mix, validate_plan) over this one, so it lacks explicit alternatives but the in-tool context is clear.

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