Shoon A2A Trust Services (pilot)

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

With no annotations, the description carries the full behavioral burden, and it does a good job: it discloses the network fetch, the report fields, the verbatim check, the sandbox/testnet context, and Ed25519-signed deliverables. It stops short of describing mutation/safety characteristics or rate limits, but the described behavior is notably transparent.

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 first states core functionality, the second lists exclusions, the third gives sandbox/pricing/signing context. The most decision-relevant information is front-loaded.

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 nested-parameter tool with no output schema and no annotations, the description covers the operation, output elements, limitations, and sandbox semantics. It does not describe the exact output structure or authentication flows, but the schema partially handles those, and the core calling context is clear.

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 only 50%: quote_id is documented in the schema but the nested input/citations fields are not. The description partially compensates by explaining that URLs are fetched and quoted text is checked, giving meaning to url and quoted_text, but it does not clarify label or the relationship between quote_id and input beyond what the schema says.

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 opens with a specific verb-resource pair ('Audit citations in a document') and explains precisely what the tool does: fetch each URL over plain HTTPS, report HTTP status/content type, and check verbatim quoted text. It also states what it does not do (judge argument quality, check paywalled content), which distinguishes it from plausible siblings like claim_verification.

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 makes the intended use clear and even gives exclusions: it does not judge argument quality or check paywalled content. It does not explicitly name sibling alternatives or give a 'use when' rule, but the scope is specific enough that an agent can infer when to select it.

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 available, the description carries the full burden and does so well. It discloses fetching over plain HTTPS, checking for quoted text, not browsing the web, marking unverifiable items as 'unverified — sandbox stub', and signing deliverables with Ed25519. It also sets expectations for testnet semantics and free sandbox usage, which helps an agent judge output trustworthiness.

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 front-loaded with the core action, and each subsequent sentence adds non-redundant constraints: HTTPS fetching, no web browsing, stub fallback, and sandbox/testnet behavior. It is compact despite covering operational and trust-related 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 tool with no output schema, the description covers the main return signals: HTTP status, quoted-text presence, and 'unverified — sandbox stub'. It does not specify the exact response structure or error formatting, but the information provided is sufficient for an agent to invoke the tool correctly and interpret basic results.

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% because `input` has no schema-level description. The prose maps claims, sources, and quoted_text to the verification behavior, which adds useful meaning. However, the description never explains the relationship between `input` and `quote_id` or that they represent alternative submission flows; that important context is left entirely to the quote_id schema description.

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 precise action: 'Check factual claims against their cited sources,' and details the exact mechanism—fetching each source URL over HTTPS, reporting HTTP status, and checking whether quoted text appears. It clearly sets a boundary by saying it does NOT browse the open web or perform independent research. It does not explicitly distinguish itself from siblings such as citation_audit, so it stops one step 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?

The description makes the operating context clear: the tool only verifies claims against user-supplied source URLs and is not appropriate for open-ended research. The 'when not' guidance is explicit, but no alternative tool is named, so it lacks the full when/alternative guidance needed for 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?

With no annotations, the description carries the full burden and does so thoroughly: it discloses plain-HTTPS fetching, per-keyword match counts and snippet context, on-demand sandbox execution, no recurring scheduler yet, testnet semantics, and Ed25519-signed deliverables. This goes well beyond the bare schema.

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 dense sentences front-load the core purpose and output, then add environment and limitation context without filler. Every sentence earns its place.

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 no-output-schema, no-annotation tool with two actions and a nested input object, the description is missing action-specific parameter requirements: it doesn't state that register needs url/label or that check consumes tripwire_id. The output summary is covered, but the calling contract is incomplete.

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 description clarifies the meaning of action, keywords, and URL, and implies tripwire_id through the register/check workflow. However, it never explains the label parameter or which fields are required for register versus check, so the agent must infer key input semantics from 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 opens with a concrete action and resource: 'Watch a public URL ... for keywords,' then names the register and check actions and what the check reports. This clearly distinguishes it from unrelated siblings like citation_audit or structured_extraction.

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 a clear workflow: register a tripwire first, then trigger a check, and notes checks are on-demand with no background scheduler yet. It doesn't explicitly describe alternatives or when-not-to-use, but no sibling is an obvious alternative, so the usage context is sufficient.

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 full responsibility, and it is transparent: it discloses plain-HTTPS fetching, the limited scope of checks, the honest labeling of results, and sandbox/testnet semantics with Ed25519-signed deliverables. This goes well beyond the schema.

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 sentences deliver the core purpose, the operational behavior, the key limitation, and relevant environment context. Every clause adds value, and no content is redundant with the schema.

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 no annotations and no output schema, the description adequately explains what the tool does, what it reports, what it cannot do, and the sandbox environment. It stops short of describing the exact response format, but the two reported outcomes—evidence resolution and artifact-text presence—are clearly stated.

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 only 50%, but the description adds meaning by explaining how evidence URLs are used and what expected_text means. quote_id is already documented in the schema, and work_claim is self-evident from the tool's purpose. Some nested fields such as label remain unexplained, preventing a 5.

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 opens with a specific verb and object: 'Audit a claim that work was performed.' It then clarifies that this is an evidence-resolution check rather than independent verification, which distinguishes it from sibling tools like claim_verification or citation_audit.

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 clear context for when to use the tool: checking evidence URLs and expected artifact text. It also provides an explicit exclusion—'It CANNOT independently verify that work happened'—so an agent knows not to use it for final verification, though it does not name sibling alternatives directly.

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 provided, the description carries the full burden and does so thoroughly: it discloses deterministic heuristics, lack of semantic understanding, labeled output, sandbox/testnet semantics, free testing, and Ed25519-signed deliverables. This is far beyond a minimal description.

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 dense sentences with no filler. The core action is front-loaded, the critical non-LLM caveat is immediate, and the sandbox/signing notes are concise and relevant.

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 no output schema and no annotations, and the description does not specify the exact output structure beyond 'plainly labeled heuristic.' The auth/payment flows are only present in the schema, and the description leaves the reader to infer output format and operational details.

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 only 50% and the description adds some meaning by mapping 'text' and 'fetched page' to the input object, but it does not detail parameter behavior. The quote_id parameter is well described in the schema itself, so the description's compensation is partial rather than complete.

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 opens with a specific verb and resource: turning messy text or a fetched page into structured data. It explicitly names the extraction targets (headings, links, tables, key statistics) and distances itself from LLM extraction, making its purpose and scope unambiguous.

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 clearly states when to use the tool (messy text or a fetched page) and what it is not for (not LLM extraction, no semantic understanding). It does not explicitly compare against sibling tools, but the exclusion is strong enough to guide selection.

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