Server Quality Checklist

Tool Scores

• kb_listB
  3.4/5.0
  Behavior2/5

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

  No annotations are provided, so the description carries the full burden. It discloses what the tool returns but does not state that it is read-only, nor any limits, errors, or side effects. More critically, the sentence 'Aucune base n'est actuellement enregistrée' asserts a current state that is not a permanent behavioral trait and may become false, misleading the agent into expecting an empty list. This adds confusion rather than transparency.

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

  Conciseness3/5

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

  The description is brief and front-loaded with the core purpose. However, the second sentence is a dynamic state claim that is not a stable fact, adding noise and potentially misleading the agent. This detracts from conciseness because the information is not consistently true and distracts from the tool's function.

  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?

  For a simple list tool with one optional parameter and no output schema, the description explains the output fields sufficiently. The parameter is documented in the schema, so the agent can discover it. There is no mention of error handling, but for a read-only list operation, this is not critical. The description covers the essential outcome and structure adequately.

  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?

  The schema description coverage is 100% and includes a detailed description of the single parameter 'include_pending_concerns' explaining its purpose and utility. The tool description adds no parameter information, but the baseline of 3 applies because the schema already covers semantics effectively.

  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 lists knowledge bases on the hub, specifying the returned fields (object, document count, pending proposals). This distinguishes it from siblings like kb_search (search) and kb_read (read a specific base). The core purpose is unambiguous, though the second sentence introduces confusion about current data state.

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

  Usage Guidelines3/5

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

  The description implies this is for getting an overview of all bases, but it does not explicitly mention when to use it versus alternatives like kb_search or kb_propose. There are no exclusions or alternative routing, leaving the agent to infer usage from the tool name and sibling context. This is implied usage, not explicit guidance.

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

• kb_governanceA
  4.1/5.0
  Behavior4/5

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

  No annotations are provided, so the description carries the disclosure burden. It names the two returned artifacts, notes the conditional existence of schema.yaml, and states that no bases are currently available. It does not cover errors or output formatting, but for a simple retrieval tool the behavioral information is sufficient.

  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 concise sentences with the core return behavior front-loaded. The final sentence about available bases is short but potentially relevant. Minor awkwardness, but no filler or redundancy.

  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?

  The tool is simple: one required parameter and no output schema. The description covers what is returned, when to use it, and current availability. Missing explicit error or response-shape details, but these are not critical for a one-parameter retrieval tool.

  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?

  The input schema already documents the single 'base' parameter at 100% coverage, including that it is the manifest `name` field. The description adds no additional parameter-level detail, so the baseline of 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 states a clear verb ('Retourne') and a specific resource: governance rules from GOVERNANCE.md and the frontmatter schema from schema.yaml. It also distinguishes the tool from siblings by tying it to the governance/proposal workflow rather than generic reading.

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

  Usage Guidelines4/5

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

  The description gives explicit when-to-use guidance: read before submitting via kb_propose and essential for the manager role. It does not name alternatives or exclusions, but the context is specific enough for an agent to decide when to call it.

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

• kb_readA
  4.2/5.0
  Behavior4/5

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

  No annotations are provided, so the description carries the behavioral disclosure burden. It usefully exposes the non-obvious behavior that large documents return a headings table unless force is used, and it explains the retry path. It does not detail the response format or possible errors, but covers the most important behavioral quirk.

  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?

  The description is three short, purposeful sentences. The main use case is front-loaded, followed by the volume caveat and the current environment note. There is no redundant or filler wording.

  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?

  Given no output schema and no annotations, the description provides enough to invoke the tool correctly: full-document vs section behavior, the large-document headings fallback, force override, and the fact that no bases are currently available. It is slightly limited by not describing response structure or error cases, but those are not essential for basic usage.

  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 parameter semantics are already fully documented in the schema. The description mostly re-contextualizes the behavior of section and force rather than adding new parameter-level meaning, matching the baseline for high schema coverage.

  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 opens with a specific verb and resource: 'Lit un document du corpus d'une base' and immediately explains both main modes: full document or a single section. This clearly distinguishes kb_read from sibling tools like kb_list and kb_search.

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

  Usage Guidelines4/5

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

  The description gives concrete usage guidance: for large documents without a section, a headings table is returned, and the agent should retry with the desired section or use force: true. It does not explicitly compare kb_read with sibling alternatives, but the intended context is clear.

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

• kb_searchA
  4.2/5.0
  Behavior4/5

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

  With no annotations provided, the description carries the burden of behavioral disclosure. It explains that the tool is a read-oriented search, explicitly states it never returns full documents, and names the follow-up tool. The 'Aucune base' note adds important environmental context. It does not discuss failure modes or limits, but these are partially covered by schema constraints.

  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 short sentences deliver purpose, return behavior, follow-up guidance, and a status caveat without redundant filler. The most actionable information is front-loaded, and every sentence earns its place.

  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?

  For a 4-parameter search tool with no output schema and no annotations, the description provides enough context: what is searched, what is returned, what is not returned, and which sibling to use next. It could name kb_list as a way to discover registered bases, but the explicit 'no base registered' warning compensates by setting expectations.

  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 75%, and the schema already documents base, query, and mode in detail. The description adds no additional parameter-level meaning beyond stating that the search targets a corpus, so it stays at baseline. The max_results parameter lacks a description, but its semantics are inferable from its name and constraints.

  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 states a specific verb and resource ('Recherche plein texte dans le corpus d'une base') and defines the return payload: paths, titles, and relevant excerpts, never full documents. It also distinguishes itself from kb_read by explicitly routing full-document reads to that sibling.

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

  Usage Guidelines4/5

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

  The description gives clear usage context: use kb_search to find relevant excerpts, then use kb_read to read a document or section. It also provides an important current-state caveat ('Aucune base n'est actuellement enregistrée') so the agent knows the search may have nothing to operate on. It does not enumerate when not to use the tool, but the guidance is sufficient.

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

• kb_proposeA
  4.4/5.0
  Behavior5/5

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

  No annotations are provided, so the description carries the full behavioral burden. It clearly discloses that proposals are not auto-integrated, are stored in proposals/pending/, and require manager review. It also discloses a v0 limitation: resolution is not consultable via MCP and requires direct git access, plus the current absence of registered bases.

  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?

  The description is front-loaded with the primary action and then adds only high-value workflow constraints. Every sentence contributes something actionable: no auto-integration, review dependency, MCP resolution limitation, pre-check alternative, and current unusability. No filler is present.

  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?

  For a proposal-submission tool with no annotations and no output schema, the description covers the lifecycle, limitations, and pre-check workflow thoroughly. The main gap is that it does not describe what response or confirmation the agent should expect after submission beyond the pending-file location. Still, the context provided is substantial enough for safe invocation.

  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 already high at 86%, so the schema carries most of the parameter meaning. The description adds context about the base lifecycle and the pending state, but does not substantially enrich individual parameter semantics beyond the schema. This matches the baseline expected when structured parameter documentation is strong.

  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 opens with a specific verb and resource: "Dépose une proposition d'ajout ou de correction dans une base." It immediately distinguishes this from the sibling read/list/search tools by stating the proposal is not integrated automatically and is queued for review. The workflow context makes the tool's submission role unambiguous.

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

  Usage Guidelines4/5

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

  The description gives concrete pre-submission guidance: use kb_list with include_pending_concerns to check for existing pending proposals. It also warns that "Aucune base n'est actuellement enregistrée," which effectively tells the agent the tool is not currently usable. It does not exhaustively state when alternatives should replace this tool, so it falls just short of a 5.

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

• kb_hub_rescanA
  4.6/5.0
  Behavior4/5

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

  No annotations are provided, so the description carries the full burden. It discloses the key behavioral trait: the rescan only affects the calling session (mono-instance scope), which is non-obvious and crucial. However, it does not mention what happens to existing discovered bases, whether the operation is idempotent, or what the return value is. The most important nuance is covered, but some secondary behavioral details are missing.

  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?

  The description is two concise sentences plus a bolded scope note. The first sentence states the action and trigger, and the second adds the scope warning. There is no fluff, and the most important content (action + when) is front-loaded. Every sentence earns its place.

  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?

  For a 0-parameter action with no output schema, the description is quite complete. It tells the agent exactly when to invoke it (after import) and discloses the critical mono-instance limitation. It omits minor details like behavior when there are no new imports or any return value, but for such a simple tool, the provided context is sufficient for correct invocation.

  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?

  The tool has 0 parameters, and the schema is empty. The description adds no parameter info because there is nothing to explain. According to the rubric, 0 params is baseline 4, and nothing is missing; the description correctly focuses on the action and scope rather than parameter details.

  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 action ('Relance la découverte des bases dans bases-dir') and the specific trigger (after importing a bundle). It is unambiguous and distinct from siblings like kb_list, kb_search, and kb_read, which are query tools; this is a rescan operation. The verb and resource are precise.

  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 says when to call this tool: 'à appeler après avoir importé un bundle' (to be called after importing a bundle), with a concrete git clone example. It also warns about the mono-instance scope, clarifying that other sessions won't see the new base unless they rescan or restart. This gives the agent a clear trigger and context, even without naming specific siblings.

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