Refusal Desk — pre-execution task-brief tribunal

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

No annotations are provided, so the description carries the full burden of behavioral disclosure, and it discloses essentially nothing. The word 'Own' vaguely implies ownership or state mutation, but whether this reads, writes, or destroys data, what permissions it needs, and what it returns are all unstated.

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?

It is short, but that brevity comes from under-specification rather than economy of expression. The single fragment is not a usable instruction and leaves the agent with no front-loaded action or outcome.

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?

With no annotations, no output schema, and no parameters, the description is the only source of information, yet it never says what the tool returns, what 'miss accounting' means operationally, or how to interpret the result. It is inadequate for an agent to invoke this tool correctly.

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?

The tool takes zero parameters, so there is nothing for the description to clarify beyond the schema; the baseline for a no-parameter tool applies. Schema coverage is 100% for the empty object, leaving no gap to compensate for.

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 fragment 'Own per-class calibration + miss accounting' hints at a calibration domain but supplies no verb explaining whether it generates a report, records a metric, or assigns ownership. Nothing here distinguishes it from siblings like check_brief or settle_outcome, so an agent cannot tell what this tool actually does.

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?

There is no statement of when to call this tool, no prerequisites, and no mention of the sibling alternatives check_brief, duplicate_lookup, or settle_outcome. Usage has to be guessed entirely from the tool name.

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

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

With no annotations, the description carries the full burden. It discloses cost and billing ("$0.01 / trial grant / Fleet quota"), which is useful and beyond the schema, but says nothing about what screening produces, whether it is read-only or records state, what failure looks like, or latency/limits. For a gate-keeping tool this is a thin behavioral picture.

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?

A single line with the action and resource front-loaded and no filler prose. The brace notation is mildly schema-like clutter and the billing parenthetical is an odd neighbor to the core purpose, but overall it is tight and scannable.

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?

There is no output schema and no annotations, so the description must shoulder the whole burden for a tool with a nested 11-field input object. It omits the screening criteria, the return shape, and any side effects, leaving an agent unable to predict what happens when it calls this.

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 0%, and the only parameter content is "{brief, standing_rules?}", which merely restates the schema's property names and required/optional status. The nested brief object has eleven fields (goal, tools, budget_usd, deadline_epoch, acceptance, etc.) that are documented nowhere, so the description adds essentially no semantic value.

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 names a verb and resource ("Screen a task brief") and pins the timing ("before execution"), so the agent knows roughly what the tool does. However, "screen" is left undefined — it never says what the brief is screened for (policy, budget, safety, duplicates), and nothing distinguishes it from siblings like calibration_report or settle_outcome.

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?

"before execution" is a genuine usage cue that tells the agent this is a pre-flight step. Beyond that there is no when-to-use/when-not, no prerequisites, and no routing to alternative tools despite three plausible siblings with overlapping concerns.

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

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

With no annotations, the description carries the full burden. It does disclose meaningful behavioral context — cost (+$0.02) and the payment rail (x402) — plus cross-requester scope, which is genuinely useful. But it omits the return format, result semantics, and any limits, so it only partially covers the load.

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?

It is short and front-loaded, which is good, but the raw '{brief}' template placeholder is an artifact rather than explanatory content, so the brevity comes at the cost of clarity. Sized fine for the sentence count but under-specified.

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 required nested-object input, no output schema, and no annotations, the description is too thin. It explains neither the structure/content of brief nor what a duplicate search returns, so an agent lacks what it needs to invoke it correctly.

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 0% on a single nested object parameter with eleven sub-fields, and the description only echoes the parameter name as the '{brief}' placeholder. It adds zero meaning about what brief must contain (goal, principal_id, tools, constraints, etc.), failing to compensate for the coverage gap.

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 (search) and resource (semantic duplicates) with a scope qualifier (cross-requester), which is fairly distinguishing. However, it does not explicitly differentiate itself from siblings like check_brief, leaving the boundary somewhat to inference.

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?

There is no when-to-use guidance, no exclusions, and no mention of alternatives such as check_brief. The description gives no signal about what conditions should prompt an agent to choose this over a sibling.

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

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

With no annotations present, the description carries the full burden, and it does disclose two real behavioral traits: the operation is idempotent, and a forecast miss incurs a $0.05 charge against a per-principal per-period cap. That is genuine mutation-cost and retry-safety context. It still omits permissions/auth requirements, side effects on the verdict record, and what a settled verdict returns.

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?

A single dense sentence that front-loads the action and immediately qualifies it with idempotency and cost. Nothing is redundant, though the bracketed parameter list is arguably filler given the schema. Efficient and well-ordered.

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 mutation tool with no annotations, no output schema, and 5 undocumented parameters, the description is thin: it never covers authorisation, the effect of settling on the verdict, error behaviour, or the meaning of the optional fields. Idempotency and pricing are valuable but not sufficient to call it complete.

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 0% across 5 parameters, so the description must compensate and does not. It only re-lists the parameter names with optionality markers ({verdict_id, outcome, failure_class?, result?, cost_usd?}), which the required list already conveys, and never explains what failure_class versus result should contain or the units/meaning of cost_usd. The inline block adds surface area without adding semantics.

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 ('Settle a verdict') with the key fields inlined, so an agent knows this finalizes a verdict record rather than reading or listing one. The sibling tools (calibration_report, check_brief, duplicate_lookup) are functionally unrelated, so no sibling differentiation is needed. The term 'verdict' is domain jargon left undefined, which keeps it short of a 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?

There is no when-to-use or when-not-to-use guidance and no named alternative — it never says under what circumstances a verdict should be settled versus looked up or reported on. Idempotency is disclosed, but that is a behavioral property, not a selection rule. Usage must be inferred entirely from the tool name.

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