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

Annotations cover the safety profile (readOnly=false, destructive=false, idempotent=false, openWorld=false), and the description adds non-obvious behavior: it writes a markdown file under data/campaigns/, returns the plan id plus shell text, and rejects unscored offers with an error that names them. It does not state what happens on repeat calls (overwrite vs. duplicate) despite idempotentHint=false, which is the one remaining gap.

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, front-loaded with what is produced, followed by the precondition and the sibling routing. Every clause carries information; 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?

No output schema exists, but the description compensates by stating the return values (plan id plus shell text) and the side effect (markdown file location). Combined with the 100%-covered input schema and annotations, an agent has everything needed 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 coverage is 100%, so both parameters are already documented in the schema; the description mostly restates that offers come from score_offer. It adds no format or matching detail beyond the schema's 'exact text match' note, so the 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?

States a specific verb (build) and resource (campaign-plan shell) and enumerates the four artifacts produced: strategy stubs, channel briefs, numbered offer registry, scoring appendix. It explicitly distinguishes itself from the sibling check_plan ('this creates the shell; check_plan later validates everything written against it').

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?

Gives an explicit precondition ('Do NOT call it before each offer has a score_offer'), names the failing behavior when violated, and routes to the alternative sibling check_plan for validation. The when-to-use and when-not-to-use conditions are both stated.

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 the safety profile (readOnly, idempotent, non-destructive), but the description adds meaningful behavior: violations come back as an error list and never a silent pass, and it states which artifacts it inspects (registry + plan file). The 'writes nothing' clause partly restates readOnlyHint, so it falls just short of a 5.

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?

Dense but front-loaded: the verification criteria come first, then usage, then sibling distinction. Every sentence contributes, though the 'read-only tool ... writes nothing' clause is slightly redundant with the annotations.

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?

No output schema exists, and the description compensates by describing the return contract (error list, never a silent pass) and the two input sources. For a validator tool this is complete enough for an agent to invoke and interpret results 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 100%, including the 'overrides plan_id' precedence, so the schema carries the parameter meaning. The description adds no further syntax, format, or default details beyond what is already documented, making 3 the appropriate 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 specific verb (verify) and resource (campaign tactics) with the exact criteria checked: offer citation and success metric. Explicitly names the sibling build_plan and contrasts its role (creates the shell vs validates on disk), so the agent can distinguish them without opening either schema.

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?

Gives a clear trigger ('after every build_plan and again after any manual edit') plus an explicit exclusion ('do NOT use it on arbitrary markdown that never went through build_plan without first scoring its offers'). Both when and when-not are covered with a named alternative workflow.

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=true, idempotentHint=true, and destructiveHint=false, so the safety profile is covered. The description adds genuinely useful non-annotation context: this is a pre-spend planning step whose outputs (targets, anti-targets, kill criteria) drive later scoring, and re-planning is gated on kill criteria. It stops short of describing how the plan is returned or how scoreable tactics are structured.

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-loads the core action and its outputs, then layers purpose, usage, and sibling routing. Dense but every clause carries routing or scope information; the parenthetical labels ('budget black holes') add slight bulk without new meaning.

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 output schema, the description usefully enumerates what the plan contains and when it is consumed, which is enough for correct invocation. It does not address the other siblings (build_plan, check_plan), leaving a minor gap in the workflow picture.

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 both parameters are documented with examples in the schema itself. The description does not add format, syntax, or constraint detail beyond what the schema 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?

States a specific verb+resource (turn a campaign objective into a pre-spend plan) and enumerates the concrete artifacts produced (offer targets, channel targets, anti-targets, kill criteria). It explicitly separates itself from score_offer, so an agent can route correctly without opening either schema.

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?

Gives explicit timing ('before score_offer on any new product/launch') and an explicit re-use condition ('re-plan only when kill criteria trigger'). It also names the alternative tool and the boundary condition that selects between them, leaving nothing to inference.

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?

Discloses a real side effect beyond the annotations: the score is written to an in-memory registry keyed by exact offer text, is lost on restart, and should be re-scored before finalizing a plan. It also explains that every dimension returns WITH its evidence (auditable, not a vibe), which the annotations do not convey.

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 purpose and scoring model, then side effect, then exclusions and sibling routing. Dense but every sentence carries decision-relevant information; 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?

Despite no output schema, the description explains the return (per-dimension evidence + grade) and flagging of statefulness (cache lost on restart). Combined with the sibling map and explicit scope limits, an agent has everything needed to call 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 coverage is 100%, so the schema already carries both parameters. The description adds value by tying 'audience' to a specific weighted dimension (audience specificity, 24 points) and stressing 'the exact sentence a prospect would read', clarifying the precision the 'offer' string must have.

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 (Score) and resource (one offer statement) plus the exact scoring model and grade bands, so the agent knows precisely what comes back. It also explicitly names what it is not (landing-page/copy critique) and distinguishes itself from plan_campaign and build_plan.

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?

Gives an explicit exclusion ('Do NOT use it as a general copy critique — it grades offer statements, not landing pages') and an explicit sibling routing map: plan_campaign plans strategy, this grades a sentence, build_plan consumes the scores. Nothing about when to reach for it is left to inference.

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