PHION Task Router

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

Annotations already declare readOnly/idempotent/non-destructive, and the description goes further with material context the annotations do not carry: the call is free, advisory, and 'never authorizes payment, moves funds, or executes a provider'. It also discloses abstention behavior. This is exactly the extra behavioral payload the description should provide.

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?

Four tightly packed sentences, with the primary routing instruction front-loaded. The first sentence's long enumeration of capability categories is functional usage guidance rather than filler, but it is on the verbose side and could be trimmed.

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 tool with a nested constraints object and no output schema, the description does cover the core contract (what it does, what it returns, non-execution guarantees, abstention). It does not explain the constraint dimensions an agent may need to narrow results, which is the one notable completeness gap.

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 50% — the `task` parameter is well documented (natural language, do not translate into a PHION service name) and the description reinforces that. But the `constraints` nested object has 14 sub-properties (privacy, maxPrice, allowedPaymentRails, etc.) with no descriptions in either the schema or the description, so the filtering vocabulary is undiscoverable. At borderline coverage, baseline 3 is appropriate, though the constraints gap is real.

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 (find the best service for a task) and additionally describes what it returns ('best evidenced options') and its advisory nature. An agent can tell exactly what this call accomplishes without opening the schema, and no siblings exist to confuse it with.

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?

'Use this first whenever an agent needs external work, data, inference, search...' gives an explicit, front-loaded trigger condition, and 'abstains when none is trustworthy' tells the agent what happens when the tool has nothing to offer. Nothing is left to inference about when to invoke it.

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