DEEPBOM

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

Annotations already declare readOnly/idempotent/non-destructive/openWorld=false, so safety is covered. The description adds genuinely new behavioral context: analysis runs in a 'browser-local analyzer' and returns a 'hash-bound static evidence summary', which tells the agent the work is local and the output is tamper-evident static evidence rather than live measurement.

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 the use trigger, followed by a behavioral note and an exclusion list. Every sentence carries distinct routing or behavioral information with 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 2-parameter tool with an output schema and full safety annotations, the description supplies the scoping, trigger, exclusions, and behavioral model an agent needs. Return format is covered by the output schema, so nothing essential is missing.

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 50%: analysis_depth is well described in the schema (enum plus when-to-use note about reading the full attachment), but the required 'file' parameter is only a $ref with no textual description. The description adds no parameter meaning at all, so it does not compensate for the uncovered file parameter; baseline 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 (inspect/analyze) and resource (attached model deployment artifact), and enumerates the exact formats accepted (TFLite, ONNX, GGUF, SafeTensors, Core ML, ExecuTorch), which sharply scopes the tool. It does not name siblings like deepbom_publish_analysis, but the action+resource pairing is concrete enough that an agent can identify it without opening the 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?

Explicitly gives the trigger ('Use when the user asks to inspect an attached ... artifact') and an extensive exclusion list (training checkpoints, measured latency, task accuracy, clinical validity, actual runtime placement). This is exactly the when/when-not guidance that lets an agent route correctly.

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 the safety profile (readOnly, idempotent, non-destructive, closed-world), so the bar is lower. The description adds meaningful context beyond them: it returns metadata only and performs no file inspection, which is exactly the behavioral distinction an agent needs here.

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?

Two tight sentences, front-loaded with the 'use before analysis when...' trigger, then the scope exclusion. No filler and nothing redundant.

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?

An output schema exists, so return values need no explanation, and the annotations carry the safety profile. Combined with the explicit usage trigger and the file-inspection exclusion, this is sufficient for an agent 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?

The tool takes zero parameters, which is the baseline-4 case; there is nothing in the schema for the description to supplement. No parameter semantics are needed or expected.

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 function (returns capability metadata) and explicitly scopes out file inspection, which cleanly separates it from deepbom_analyze_file. It is clear what the tool returns, though the 'what it does' is partly carried by the name/title rather than a strong verb+resource phrasing.

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?

It gives explicit trigger conditions: use it before analysis when format support, privacy boundaries, or the ChatGPT-vs-local path distinction is unclear. That is a concrete when-to-use rule, and the 'does not inspect a file' clause implicitly steers file work to the analyze 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?

Annotations declare readOnlyHint, idempotentHint, non-destructive, and non-open-world. The description adds that the tool 'validates its structure' and accepts 'analysis results, never model bytes', which is useful behavioral context. However, it doesn't explain what happens upon validation failure, retry behavior, or the output beyond what the output schema covers.

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?

Two concise sentences, front-loaded with the core action. No wasted words. Could be slightly more informative but is efficient.

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 the complex nested input, rich annotations, and presence of an output schema, the description covers the basic scope and clarifies it's not for model bytes. However, it doesn't address common failure modes, validation expectations, or how the result is used downstream. The parameter semantics gap is the main shortfall.

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%, so the description carries the full burden for the single complex 'result' parameter with many nested required fields. The description only states it accepts 'one bounded browser-produced DEEPBOM result' and 'never model bytes', which gives some scope but doesn't explain the required schema fields or the implications of the nested structure. This leaves significant ambiguity for an agent constructing the argument.

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 and resource: 'receives', 'validates', 'returns' a 'DEEPBOM analysis result' to the conversation. This is clear enough to distinguish from deepbom_publish_error (publishes errors rather than analysis results) and deepbom_analyze_file (produces analysis rather than returns it). It could more explicitly name the sibling it's not.

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 ('browser-produced DEEPBOM result', 'returns it to the conversation') but does not state when to use this tool versus alternatives or when not to use it. There's no explicit routing guidance for the agent.

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, destructiveHint=false, and openWorldHint=false, so the agent knows this is a safe, local, repeatable read operation. The description adds the important constraint that it accepts diagnostic text and never model bytes, which is useful context beyond the annotations. However, it doesn't describe what the recovery action looks like or how errors are processed.

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?

Two concise sentences that front-load the purpose and include an important constraint. No waste, though it could be slightly more informative without becoming verbose.

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 has a detailed input schema and an output schema, so the description needn't explain return values. However, for an error-handling tool with no usage guidelines and no clarification of how the recovery action is generated or used, the description leaves gaps in contextual understanding.

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%, but the nested error object has eight required fields with const values and additionalProperties=false, which provides strong structural guidance. The description only clarifies that the input is diagnostic text, not model bytes, which adds marginal semantic value. Baseline 3 is appropriate given the rich schema structure compensates for the lack of text descriptions.

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: receives a bounded browser-produced DEEPBOM error and returns a recovery action to the conversation. Clear purpose, but does not differentiate from sibling tools like deepbom_publish_analysis or deepbom_capabilities beyond the word 'error'.

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?

No guidance on when to use this tool versus alternatives like deepbom_analyze_file or deepbom_publish_analysis. The description implies it's for error recovery but doesn't state the triggering condition or when it should not be used.

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