GPT-5.6-luna Agent

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

Annotations already indicate readOnlyHint and idempotentHint. The description adds that it does not run the model and does not require a bearer token, providing additional behavioral transparency beyond 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 two sentences, efficiently conveying the returned items and key behavioral traits. No extraneous content.

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?

Given no parameters and sufficient annotations, the description covers the essential return values and behavioral notes. No gaps identified.

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 has zero parameters, so the description does not need to elaborate on parameter semantics. Baseline score of 4 applies per instructions.

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 explicitly states the tool returns agent identity, per-task price, MCP endpoint, and access URL. It clarifies this is a metadata call that does not run the model. The verb 'returns' is clear, and it distinguishes from sibling tools that deal with runs.

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 notes that the call does not run the model and does not require a bearer token, implying it should be used when metadata is needed without model execution. While it does not explicitly mention when not to use, the context of sibling tools for runs provides clear differentiation.

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?

Annotations already declare destructiveHint=true and idempotentHint=true, covering the safety profile. The description adds a meaningful constraint—only queued or running runs can be cancelled—but does not disclose what happens after cancellation (e.g., irreversible, async, effect on results). This adds some value but not rich behavioral context.

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, front-loaded sentence with zero wasted words. It conveys the action, resource, and constraint efficiently.

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 one-parameter, destructive, idempotent tool, the schema and annotations cover most of what an agent needs. The main gap is that no output schema exists and the description does not mention what the tool returns after a successful cancel or how errors (e.g., run already finished) are surfaced, which would be useful but not critical.

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 input schema covers the sole parameter run_id with 100% description accuracy ('AgentFarm run identifier'). The tool description adds no additional parameter meaning beyond what the schema already provides, so the baseline of 3 applies.

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 states a specific verb ('Cancels'), a resource ('AgentFarm run'), and a scope qualifier ('authenticated queued or running'). This clearly distinguishes it from sibling tools like create_run, get_run_status, and get_run_result, so an agent knows exactly what the tool 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?

The description implies this tool is only for queued or running runs, but it does not explicitly say when to use it versus alternatives or mention any exclusions. Context about siblings is present in the signal data but not referenced in the description itself.

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?

Beyond the annotations (readOnlyHint=false, idempotentHint=false), the description discloses that the task is asynchronous ('Queues'), requires authentication ('authenticated'), and has cost implications ('can consume paid model capacity'). It also reveals the lifecycle behavior by routing to polling and result tools. No contradiction with annotations — the mutation/async nature aligns with readOnlyHint=false. It doesn't cover failure modes or rate limits, but the annotation bar is satisfied.

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?

Three sentences, each earning its place: the core action, the cost warning, and the follow-up workflow. The most important information (what the tool does) is front-loaded, and there is zero filler or repetition of schema content.

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?

The tool is simple (3 fully documented params) and the workflow chain is stated, but there is no output schema and the description never says what the create call returns (e.g., a run ID to pass to agentfarm_get_run_status). An agent must infer the response shape to chain the lifecycle correctly. Cost and async caveats are covered, but the missing return-value semantics is a real 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 100% — all three parameters (task, task_input, idempotency_key) already have meaningful descriptions in the schema, including the idempotency behavior of idempotency_key ('the same key and input return the original run'). The tool description itself adds no parameter-level detail, placing it at the baseline 3 for high-coverage schemas.

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 uses a specific verb ('Queues') with a clear resource ('authenticated asynchronous task'), and the title anchors it as 'Create a GPT-5.6-luna Agent run.' It distinguishes itself from siblings by naming the downstream workflow (agentfarm_get_run_status, agentfarm_get_run_result), making it unambiguous that this is the creation step rather than a status or result read.

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 gives explicit workflow guidance: 'Poll with agentfarm_get_run_status, then read agentfarm_get_run_result,' telling an agent exactly which sibling tools follow this one. The cost warning ('This can consume paid model capacity') implies agents should invoke cautiously. It lacks explicit when-not-to-use conditions or mention of the cancel/agent-info siblings, so it stops short of a 5.

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?

Annotations already declare readOnlyHint, idempotentHint, and non-destructive nature. The description adds context that the tool returns final summaries and URLs for authenticated runs, which is useful beyond the structured fields.

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?

Single sentence, front-loaded with action and result, contains no unnecessary words or details.

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 simple tool with one parameter and no output schema, the description is nearly complete. Could be improved by explicitly stating the run must be finished, but the sibling context suffices.

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 coverage is 100%, and the schema already describes the 'run_id' parameter. The tool description adds no additional meaning to the parameter beyond what's in the schema.

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 uses a specific verb ('Returns') and identifies the resource ('final summary and artifact download URLs'), clearly distinguishing it from sibling tools like agentfarm_get_run_status (which returns status) and agentfarm_create_run (which creates runs).

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 implies usage for completed runs ('final summary') but does not explicitly state when to use this tool versus alternatives like agentfarm_get_run_status or what constitutes an 'authenticated run' condition.

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?

Annotations already cover safety (read-only, idempotent, non-destructive). The description adds value by specifying the returned status values and mentioning authentication context, though it does not detail failure modes or rate limits.

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, concise sentence that immediately conveys the tool's function with no unnecessary words.

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?

The description is nearly complete for a simple status check tool, listing the possible status values. However, it does not describe the exact return structure (though no output schema exists), so there is a minor 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 coverage is 100%, so the schema already documents the parameter well. The description does not add new semantic meaning beyond what the schema provides.

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 uses a specific verb ('Returns') and resource ('status'), enumerates possible values, and implicitly distinguishes from sibling tools that handle agent info, creation, and results.

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 a clear context for use (checking run status) but does not explicitly state when to avoid this tool or recommend alternatives among siblings.

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