Server Quality Checklist

Tool Scores

• read_urlA
  4.3/5.0
  Behavior4/5

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

  Annotations already declare readOnlyHint=true and openWorldHint=true, so the safety profile is covered. The description adds meaningful behavioral context beyond annotations: hidden-text stripping with threat reporting, stealth-browser fetching for blocked sites, and per-domain rate limiting. It does not contradict annotations. The only minor gap is no mention of failure modes or what 'grounded content with provenance' structurally looks like, though the output schema likely covers return shape.

  Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

  Conciseness4/5

  Is the description appropriately sized, front-loaded, and free of redundancy?

  Three tight paragraphs, each earning its place: first defines the action and output, second explains the hidden-text safety behavior and value proposition, third gives usage guidance, alternatives, and rate-limit caveat. It's appropriately sized for a tool with 5 parameters and meaningful behavioral nuance, though it leans slightly long relative to its sibling-differentiation needs.

  Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

  Completeness4/5

  Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

  With an output schema present, return-value details are handled structurally. The description covers the tool's distinctive behaviors (stealth fetching, hidden-text stripping, provenance), usage guidance, rate limits, and alternatives. For a tool of moderate complexity with rich schema and annotations, the description is complete enough. It could add slightly more on error/failure behavior, but it is not a material gap.

  Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

  Parameters3/5

  Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

  Schema description coverage is 100%, so all five parameters are documented in the schema. The description resists repeating schema detail and instead adds cross-parameter context (e.g., 'reported in `threats`' ties include_hidden and hidden-text stripping together). The `query` and `format` behaviors become clearer through the description's grounding framing, but the schema already does the heavy lifting. Baseline 3 is appropriate.

  Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

  Purpose5/5

  Does the description clearly state what the tool does and how it differs from similar tools?

  The description uses a specific verb+resource ('Fetch one web page through the stealth browser') and clearly states what it returns ('clean, grounded content with provenance'). It explicitly differentiates from siblings: 'use `search` to find URLs, or `research` to search and read in one call', naming the alternatives directly. The reference to 'sites that block plain fetchers' adds a clear use-case distinction.

  Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

  Usage Guidelines5/5

  Does the description explain when to use this tool, when not to, or what alternatives exist?

  The description explicitly states when to use this tool vs alternatives: 'Reads a URL you already have: use search to find URLs, or research to search and read in one call.' It also discloses the rate-limiting behavior ('5s apart by default'), setting expectations about latency. This is explicit when/alternative guidance with concrete operational context.

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

• searchA
  4.5/5.0
  Behavior4/5

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

  Annotations declare readOnlyHint=true and openWorldHint=true, which aligns with the description (read-only, returns links, nothing fetched). The description adds valuable context beyond annotations: explains the dual backend (SearXNG when SEARXNG_URL set, otherwise stealth-browser rendered page) and clarifies no content is fetched until requested. Confirms the lazy-fetch behavior without contradiction.

  Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

  Conciseness5/5

  Is the description appropriately sized, front-loaded, and free of redundancy?

  Well-structured with a clear first line stating the core purpose, followed by a 'Not for' exclusion line, a usage-recommendation paragraph, and a backend-detail sentence. Every sentence earns its place; nothing is redundant or padded. Approximately 90 words for a tool that does meaningful work.

  Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

  Completeness5/5

  Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

  Given the output schema exists, return values are covered structurally. The description covers purpose, exclusions, alternative tools, workflow recommendation, and backend behavior. For a 2-param read-only search tool, this is fully complete—the agent has everything needed to select and invoke correctly.

  Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

  Parameters3/5

  Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

  Schema coverage is 100% and both parameters (query, limit) are described in the schema itself. The description mentions the tool returns ranked hits but adds nothing about parameter syntax or semantics beyond schema. Baseline 3 is appropriate given full schema coverage; description adds minimal extra parameter value but the schema fully documents them.

  Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

  Purpose5/5

  Does the description clearly state what the tool does and how it differs from similar tools?

  Clear verb+resource ('Search the web') with explicit output format (ranked hits: title, url, snippet, engine). Strongly distinguishes from siblings by explicitly stating what it is NOT for (returning page content vs links).

  Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

  Usage Guidelines5/5

  Does the description explain when to use this tool, when not to, or what alternatives exist?

  Excellent when/when-not guidance: explicitly says use read_url for one known URL, research when you want the answer rather than links. Also gives a concrete workflow pattern: find pages, then pass URLs to read_url. This is exactly the differentiation agents need among the three sibling search-ish tools.

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

• researchA
  4.7/5.0
  Behavior4/5

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

  Annotations already declare readOnlyHint=true and openWorldHint=true, and the description adds meaningful behavioral context: that it reads through a stealth browser, returns provenance receipts, strips injection payloads, and reports failed sources in `sources` rather than failing the whole call. This adds real behavioral value beyond the annotations, though return format specifics could be further detailed beyond what the output schema shows.

  Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

  Conciseness5/5

  Is the description appropriately sized, front-loaded, and free of redundancy?

  Three paragraphs, each earning its place: what it does, how failures are handled, and when to use alternatives. No wasted words, well front-loaded with the purpose sentence.

  Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

  Completeness5/5

  Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

  For a complex tool with an output schema and good annotations, the description covers usage differentiation, failure semantics, cost/latency control, and behavioral traits. An output schema exists so return values needn't be spelled out. Highly complete for its complexity.

  Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

  Parameters4/5

  Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

  Schema coverage is 100%, so the schema documents all three parameters. The description adds value by explaining max_sources is the main cost/latency control and that values outside 1-10 are clamped, and that max_tokens has a server default (GROUNDHOG_MAX_TOKENS). It supplements but doesn't fully compensate the structured schema since the schema is already rich.

  Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

  Purpose5/5

  Does the description clearly state what the tool does and how it differs from similar tools?

  States a specific verb (research/search) plus resource (the web), and clearly differentiates from siblings by noting this combines `search` + repeated `read_url` into one call. The distinction from read_url and search is explicit.

  Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

  Usage Guidelines5/5

  Does the description explain when to use this tool, when not to, or what alternatives exist?

  Gives explicit when-to-use guidance: prefer `read_url` when you already have the URL, `search` when you only want links. Also warns this is the slowest of the three due to rate limiting per domain, and should be avoided for a single known page. This is model-tier guidance naming alternatives and exclusions.

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

• statusA
  4.7/5.0
  Behavior4/5

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

  The description names the specific output field `browser_reachable` and the `hint` field, giving the agent concrete knowledge of what to expect. Annotations (readOnlyHint=true) already signal this is a safe read operation. The description adds value by explaining what the output means behaviorally (telling the agent what to do when unreachable), going beyond the annotation safety profile.

  Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

  Conciseness4/5

  Is the description appropriately sized, front-loaded, and free of redundancy?

  Two sentences, front-loaded with the core purpose first, followed by actionable diagnostics. No wasted words. Slight deduction could be argued for not being maximally compact, but the conditional instruction earns its place since it directly guides agent behavior. Efficient and purposeful.

  Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

  Completeness5/5

  Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

  This is a simple, parameterless diagnostic tool with an output schema present, so the description doesn't need to explain return values extensively. The description covers the purpose, when to use it, and the conditional action to take based on the result. For a tool of this complexity, this is complete — nothing meaningful is missing.

  Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

  Parameters5/5

  Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

  The tool has 0 parameters, so there is no parameter meaning to convey — the schema itself documents this fully (100% coverage by definition since there are no params). The task is parameter-free, so a baseline of 4-5 applies, and the description fully compensates by explaining what the tool checks and what its output communicates.

  Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

  Purpose5/5

  Does the description clearly state what the tool does and how it differs from similar tools?

  The description clearly states the tool's purpose: checking whether Groundhog can reach the stealth browser. It uses a specific verb ('check') with a clear resource ('reach the stealth browser') and includes a diagnostic framing, distinguishing it from sibling tools (read_url, research, search) which are substantive data-gathering operations rather than setup diagnostics.

  Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

  Usage Guidelines5/5

  Does the description explain when to use this tool, when not to, or what alternatives exist?

  The description gives explicit when-to-use guidance: 'Call this to diagnose setup before fetching.' It also provides conditional follow-up instructions: if `browser_reachable` is false, follow `hint` to start the browser, then retry. This is directly actionable and tells the agent exactly the workflow to follow.

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