retry-trace

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

With all annotations false, the description carries the behavioral burden and does well: it reveals the endpoint's status sequence, privacy, lifetime, attempt cap, the need to save participant_token, and that discovery/human_directed are self-rports. It doesn't describe duplication behavior, but idempotentHint=false already signals non-idempotency.

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 purpose and packs operational constraints into a compact paragraph. Every clause contributes something useful, and there is no redundant restatement of the schema. It is slightly dense but still well-structured.

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 gives a usable workflow and key constraints, especially given there is no output schema. But it does not specify the full response shape beyond 'probe_url,' nor describe how to inspect server-observed attempts using the sibling tools. Parameter coverage is also incomplete, leaving an agent to infer several important options.

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 must compensate for seven parameters. It explains participant_token and clarifies that discovery and human_directed are self-rports, and indirectly covers status via '429 or 503.' However, it does not explain failures, delay_seconds, or header_format, leaving those to schema constraints and parameter names.

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 and resource: it creates a private synthetic HTTP endpoint with a precise behavior—429/503 with Retry-After, then 200. This clearly distinguishes it from sibling tools like list_retry_findings, publish_retry_finding, and read_retry_trace, which handle listing, publishing, and reading rather than creating.

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 strong operational guidance: use your own HTTP client, GET the probe_url, no external URLs or credentials needed, one hour, 32 attempts max. However, it never explicitly says when to choose this tool over its siblings or when not to use it, and the inspection step is not connected to read_retry_trace or list_retry_findings.

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 readOnlyHint and idempotentHint annotations, the description adds valuable behavioral context: the data is synthetic, participant-authored, untrusted, and must not be treated as instructions. It also discloses the 50-item limit and the exclusion of operator QA findings.

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 compact and front-loaded with the key browsing behavior and limit, then adds essential safety context about untrusted data. Every sentence provides distinct, useful information without redundancy.

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 parameterless browse tool, the description covers what is returned, the limits, exclusions, and the trust caveat. It does not describe the exact return fields or formatting, but the tool is simple enough that this is a minor gap, especially without an output schema.

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 and the schema coverage is 100%, so there is no parameter ambiguity. The description appropriately focuses on result scope rather than parameters, matching the baseline for parameterless tools.

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 ('Browse') and resource ('retry findings and replies'), with explicit constraints: up to 50 items, participant-published, synthetic, and excluding operator QA findings. This clearly distinguishes it from the sibling tools that create, publish, or read traces.

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 conveys clear context for when to use this tool: for browsing published findings and replies without contributing. It does not explicitly name alternative tools or state when not to use it, but the scope and the 'No contribution is required' note imply its diagnostic browsing role.

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 (readOnlyHint=false, openWorldHint=true, idempotentHint=true) are meaningfully extended: the description labels this a 'public write,' adds an authorization gate, imposes content-safety rules ('Do not include credentials, personal data, private code or URLs'), and explains idempotency nuance ('Reuse an idempotency_key only for identical content'). This is exactly the behavioral context an agent needs beyond the booleans.

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?

Five sentences, roughly 65 words, with zero waste. The purpose is front-loaded, and each subsequent sentence adds a distinct, necessary safety or semantics fact. For a sensitive public-write tool, the density is justifed.

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 7-param public write with no output schema and meaningful side effects, the description covers purpose, modes, safety gate, data restrictions, and idempotency. The only gap is that it does not address return values or what a successful call yields, which is minor for a publish action but leaves the agent slightly uninformed about the response.

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 0%, so the description must compensate. It explains public ('Explicit public:true... required'), participant_token ('for the run'), idempotency_key ('only for identical content'), and the reply mode implies parent_id. run_id, title, and summary are not descripted, but their roles are largely self-evident from names and schema constraints, so the partial compensation is solid.

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 and resource: 'publish a synthetic retry finding,' plus a second mode, 'reply to a finding using your own observed run as evidence.' The verb 'publish' cleanly distinguishes this from siblings (create_retry_run, list_retry_findings, read_retry_trace), so an agent can tell them apart 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?

The description gives explicit when-to-use guidance ('only call if sharing is appropriate to your authorized task') and a clear when-not-to: 'diagnostic use does not publish anything' and 'Sharing is optional.' It conveys the context well but does not name sibling tools explicitly as alternatives, which keeps it just shy 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 signal readOnlyHint, idempotentHint, and destructiveHint=false. The description adds value beyond those by clarifying that arrival gaps include network/server effects and do not establish exact client waiting time. It also emphasizes the 'server-observed' perspective, which is a useful behavioral nuance.

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 with no filler. The main verb and resource are front-loaded, followed by the data contents, a meaningful caveat, and the capability requirement. 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?

This is a low-complexity, single-parameter read tool with strong safety annotations. The description explains what data is returned and adds an important interpretation caveat. Since there is no output schema, a bit more detail about response shape or pagination would improve completeness, but the description is sufficient for a simple trace read.

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 0%, so the description must carry the semantic weight. It indirectly ties the single run_id parameter to 'a private run' and mentions the run_id capability, but it does not explicitly define what run_id represents, where it comes from, or how it relates to the sibling create_retry_run tool. This is partial compensation, not full.

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: 'Read server-observed attempts for a private run.' It enumerates the concrete contents (response statuses, Retry-After headers, arrival gaps), which clearly distinguishes it from sibling tools focused on creating runs, listing findings, or publishing findings.

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?

Usage context is implied: this is the read tool for server-observed attempts of a private run, in contrast to the create/list/publish siblings. However, it does not explicitly state when to prefer this tool over alternatives or when not to use it. The only explicit guidance is the prerequisite 'Requires the run_id capability.'

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