Server Quality Checklist

Tool Scores

• scheduleA
  3.5/5.0
  Behavior3/5

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

  Annotations already establish that this is a non-read-only, non-idempotent operation, and the description adds the scheduling-delivery context. It does not describe side effects like cancellation removing a queued item, delivery timing tolerance, or idempotency nuances, though some of these are captured in the input schema's parameter descriptions.

  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?

  The description is a single front-loaded sentence with no filler. It clearly states the core behavior first and uses a short second clause to cover management actions. It is appropriately concise, though 'manage' is a somewhat broad term that could be more specific.

  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 the rich input schema that fully documents the create/list/cancel actions and their parameters, the high-level description is sufficient for an agent to orient itself. The description does not mention return values, but no output schema exists and the action descriptions in the schema already indicate what list returns.

  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 the input schema already documents all 16 parameters and the action enum. The tool description itself adds no parameter-level meaning, but it is not required to compensate because the schema is fully explicit.

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

  Purpose4/5

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

  The description uses an active verb and a clear resource: 'Queue mail for later delivery, and manage what is queued.' This makes the tool's main purpose understandable and distinct from immediate send tools, though it does not explicitly name sibling tools or explain how it differs from email_compose or draft.

  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 phrase 'for later delivery' implies when the tool should be used, and 'manage what is queued' covers listing and canceling scheduled sends. However, there is no explicit guidance about when not to use it, no mention of alternatives like email_compose for immediate sending, and no stated prerequisites for scheduling.

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

• email_readA
  3.8/5.0
  Behavior4/5

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

  Annotations already declare readOnlyHint=true, openWorldHint=true, idempotentHint=true, and destructiveHint=false, covering the safety profile. The description adds genuinely non-obvious behavior beyond that: single-page returns with has_more/next_offet continuation ('Only has_more: false means you have seen everything') and windowed bodies via body_truncated/body_next_offet. No contradiction with annotations.

  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?

  Four short sentences with zero filler. The purpose is front-loaded in sentence one, and each of the two protocol traps (page continuation, body windowing) gets one dense sentence that earns its place. Nothing is redundant with the schema.

  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 30-parameter, 7-action tool with no output schema, the description captures the trickiest protocol contract — pagination and body windowing — that an agent would most likely get wrong. Operational details like per-action semantics, the 10 MB attachment budget, inbox disambiguation, and provider quirks are carried by the thoroughly described schema, so the division of labour is reasonable. Return shapes aren't covered, but the absent output schema makes that a minor 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%, and the parameter docs already explain the continuation semantics (offset: 'Pass the previous response's next_offset exactly'; body_offset: 'Pass back body_next_offset to continue a truncated body'). The description reinforces this at the protocol level but adds little the schema didn't already say, so the baseline 3 applies.

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

  Purpose4/5

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

  The description states a specific scope: 'Read, list and search email in one inbox.' The verbs (read, list, search) name the concrete operations and the resource is unambiguous. It doesn't explicitly differentiate from siblings like inbox_list or email_organize, but the operation set is precise enough that an agent can tell what this tool does at a glance.

  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 gives no explicit when-to-use-vs-alternatives guidance for sibling tools. It does teach the pagination loop within the tool, and the schema's `action` parameter description routes the agent to the right action (list/read/read_batch/search/attachment/extract/original) with per-action semantics. This is implied rather than explicit guidance, so a 3 fits.

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

• folderA
  3.8/5.0
  Behavior4/5

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

  Given the annotations already mark it non-read-only and destructive, the description adds useful behavioral context: Gmail labels are returned/managed instead of folders, and action-specific permission scopes are disclosed. No contradiction with annotations.

  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?

  A single dense sentence that front-loads the main purpose and then adds two high-value clarifications (Gmail label equivalence and permission split). No wasted words.

  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 description plus a rich schema covers the tool's four actions, required parameters, provider-specific behavior, and permissions. It omits any mention of alternatives, but that gap is captured under usage_guidelines; overall context is sufficient for a competent agent.

  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% and the schema already explains each parameter, including the Gmil vs IMAP interpretation of folder_id. The tool description adds only the cross-provider framing, which helps conceptual understanding but does not significantly extend the schema's param-level documentation.

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

  Purpose4/5

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

  Description states a clear verb-resource pair: 'Manage mailbox folders', and adds the critical nuance that on Gmil these are labels. It does not explicitly differentiate itself from siblings like email_organize, but the core purpose is unambiguous.

  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?

  Usage is implied rather than explicitly scoped: an agent can infer to use this tool when managing folders/labels. It provides permission guidance ('list needs read:email, the rest manage:folders') but does not name alternatives or state when NOT to use it.

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

• email_deleteA
  3.9/5.0
  Behavior5/5

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

  Beyond the destructiveHint annotation, the description discloses that mail goes to Trash and is recoverable unless permanent:true is passed, introduces the irreversible case, and calls out that the MCP client should ask for confirmation. It adds meaningful behavioral context beyond the annotations and does not contradict them.

  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 concise sentences: the first front-loads the action and scope, the second addresses safety/confirmation, and the third covers recovery, irreversibility, and auth. Every sentence earns its place with zero fluff.

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

  Completeness3/5

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

  Given a complex 21-parameter conditional schema with three action modes, the description does not mention delete, delete_batch, or search_and_delete, nor the filter-based search behavior. The schema compensates heavily, but the description alone would be incomplete for an agent deciding how to invoke this destructive multi-mode 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?

  Schema coverage is 100% with rich descriptions on every parameter, including action modes, filters, and permanent. The description itself adds no parameter-level meaning beyond highlighting permanent:true, which the schema already documents. Baseline 3 is appropriate.

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

  Purpose4/5

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

  The description states a clear verb and resource: 'Delete messages in one inbox.' It is unambiguous about the core purpose. However, it does not explicitly differentiate from siblings like email_organize or bulk_cancel, leaving the agent to rely on the name for disambiguation.

  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 usage for deleting messages and notes the required delete:email scope, but gives no explicit when-to-use guidance, exclusions, or alternatives among sibling tools. It also does not explain when to choose delete vs delete_batch vs search_and_delete—that is left to the schema.

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

• draftA
  4.1/5.0
  Behavior5/5

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

  The annotations only indicate readOnly=false and destructiveHint=true, so the description carries real weight here. It adds non-obvious behavioral detail: IMAP draft_ids change on every update, signatures are embedded at create/update time, send transmits the stored body as-is to avoid double signatures, and reply/send require specific scopes. No contradiction with annotations exists.

  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 four sentences, front-loaded with the core purpose, followed by the most important operational caveats. Every sentence adds distinct information with no filler or repetition.

  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 15 parameters and 6 actions, the schema carries most structural detail and the description supplies the missing operational context: draft_id freshness, signature behavior, and permission requirements. It is not a 5 because it does not enumerate the possible actions or mention return-value expectations, though the lack of an output schema lowers that burden.

  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 description coverage is 100%, so the baseline is 3. The description adds value beyond the schema by clarifying that include_signature is embedding on create/update and that send uses the stored body verbatim, which prevents a common double-signature mistake. It does not need to repeat every parameter because the schema already documents them thoroughly.

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

  Purpose4/5

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

  The description states that the tool manages unsent drafts and the action enum spells out six concrete operations (list, create, reply, update, send, delete), so an agent can identify its purpose. It stops short of 5 because 'Manage' is a generic verb and no explicit contrast with sibling tools such as email_compose is provided.

  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 resource 'unsent drafts' implies the main use case, and the schema action descriptions give within-tool guidance for choosing list versus create versus reply, etc. However, the description never names alternative sibling tools or states when not to use this tool, so the guidance is implied rather than explicit.

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

• approval_scheduleA
  4.2/5.0
  Behavior4/5

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

  Beyond the annotations, the description explains that this action does not queue anything immediately and only determines when an approved message is delivered. It also clarifies the future timestamp requirement. This adds useful behavioral context consistent with the idempotentHint and non-destructive annotations.

  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 tight, front-loaded sentences with no filler. The core action, key behavioral caveat, and constraint are all stated efficiently.

  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 two-parameter tool with no output schema, the description covers the essential operational context: it applies to pending approval sends, does not queue until approval, and requires a future send_at. It could mention what happens on repeated calls or error conditions, but the low complexity makes this adequate.

  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 both parameters are already documented. The description reiterates the future-time requirement for send_at and references approval_id implicitly through 'waiting for human approval', but adds little beyond the schema.

  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 ('Set') and a precise resource ('delivery time of a send that is waiting for human approval'), and further clarifies that it only schedules an approved message rather than queuing it. This clearly differentiates it from siblings like schedule, approval_decide, and approval_review.

  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 context: the tool is for sends waiting for human approval, and it explicitly notes that nothing is queued until approval. It also states the future-time constraint. It does not explicitly name alternative tools or say when not to use it, but the context is sufficient for most agents.

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

• email_composeA
  4.2/5.0
  Behavior4/5

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

  The annotations already convey this is a mutating, non-idempotent operation. The description adds a meaningful behavioral detail beyond the schema: the inbox signature is appended automatically, and it is placed above quoted text on replies and forwards. This helps the agent predict side effects and know when to set include_signature to false.

  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?

  Two compact, front-loaded sentences with no filler. The first sentence states the core operations; the second provides the one behavioral caveat an agent most needs before composing replies or forwards.

  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 tool with 17 parameters and conditional schemas, the description gives the essential high-level shape while the input schema covers details like inbox selection, attachments, idempotency, and recipient limits. The only notable omission is any mention of return values, but no output schema is provided and the tool's success/failure behavior is largely inferable.

  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 the schema already documents all 17 parameters thoroughly. The description adds only a small nuance about signature placement relative to quoted text; it does not substantially enrich parameter understanding beyond the schema baseline.

  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 names specific actions ('Send new mail, reply, or forward') and a clear resource ('from one inbox'). These verbs distinguish it sharply from sibling tools like email_read, email_delete, and email_organize, so an agent can select 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.

  Usage Guidelines4/5

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

  The description clearly establishes the use case: any time the user needs to send, reply to, or forward email. It does not list alternatives or explicitly say when not to use this tool, but the context is unambiguous enough and clearly separated from the sibling read/delete/organize operations.

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

• email_organizeA
  4.2/5.0
  Behavior4/5

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

  Annotations already mark the tool as mutating, and the description adds useful behavioral context: Gmail's move label semantics, the permission prerequisite, and the one-inbox scope. This is real operational detail beyond what the annotations convey.

  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 dense sentences cover scope, workflow prerequisite, provider behavior, permission, and sibling routing. There is no filler, repetition, or restatement of the tool name.

  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 complex tool with seven action variants and no output schema, the description provides the essential invocation context: where ids come from, what permission is needed, how Gmail behaves, and the deletion boundary. The schema handles per-action parameter constraints, so the description is appropriately balanced.

  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 the description does not need to document the 22 parameters. The prose reinforces the message-id workflow but does not add parameter-level meaning beyond what the schema already provides.

  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 starts with a clear verb+resource pairing: 'Move, copy, flag or archive messages in one inbox.' It also distinguishes the tool from email_delete by explicitly stating that deleting is a separate tool, so the agent can identify the tool's scope 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.

  Usage Guidelines4/5

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

  Gives concrete workflow guidance: retrieve message ids from email_read first, requires manage:folders permission, and explicitly routes deletion to email_delete. It does not spell out when the search-based action should be used instead of the id-based actions, but the schema's action parameter fills that gap.

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

• signatureA
  4.2/5.0
  Behavior4/5

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

  Annotations already signal non-read-only, idempotent, and non-destructive traits. The description adds non-obvious behavioral details: the signature is appended on send/reply/forward/draft/scheduled mail, and setting it overrides Gmail auto-import by marking the source 'manual'. These are useful disclosures beyond what annotations provide.

  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 concise sentences with no filler. It front-loads the core operation, then adds only high-value behavioral and permission context, earning each sentence's 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?

  The description covers authentication needs, side effects, and the server's signature behavior, while the detailed input schema covers parameter semantics, conditional constraints, and defaults. No output schema exists, but the 'action' parameter already documents what 'get' returns, so the definition is sufficiently complete 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.

  Parameters3/5

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

  Schema description coverage is 100%, and each parameter already has a meaningful description including enum values, defaults, and conditional constraints. The tool description does not need to add per-parameter semantics, so the baseline of 3 applies.

  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: 'Read or set an inbox's signature', and adds clarifying behavioral context about server-side appending. This distinguishes it from sibling tools like email_read, email_compose, or draft, none of which manage signatures.

  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 operation-specific context: 'get' needs read:email, 'set' needs send:email, and setting marks the source as 'manual', overriding Gmail auto-import. It doesn't explicitly name alternative tools, but no sibling appears to overlap with signature management, so the guidance is clear enough.

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

• bulk_cancelA
  4.3/5.0
  Behavior5/5

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

  The description adds substantial behavior beyond annotations: it states that nothing in the mailbox changes, that previews auto-expire after 15 minutes, and that cancelling records a decision instead of leaving it implicit. These details complement the idempotentHint and destructiveHint annotations and clarify that readOnlyHint=false is consistent with recording a decision.

  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 concise, front-loaded with the core purpose, and every sentence provides value: the consequence of cancellation, the single argument, the lack of side effects, and the expiry/recording nuance. No wasted words or redundant boilerplate.

  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 tool with a single parameter, full schema coverage, and annotations that cover idempotency and destructiveness, the description fully specifies the scenario, the side-effect behavior, and the operational nuance. An agent has all the information needed to invoke it correctly; no output schema is necessary for such a simple declarative action.

  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 provides 100% coverage for the only parameter, plan_id, with a description ('The plan_id returned with a bulk preview') that is clear and self-sufficient. The description merely reinforces 'Takes only the plan_id,' adding singularity emphasis but no fundamentally new semantic meaning.

  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 ('Decline'/'cancel') on a specific resource ('a previewed bulk delete or move') and the outcome ('so it can never run'). It distinguishes itself from sibling tools like bulk_execute without naming it, by describing the opposite action.

  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?

  Usage is implied: this is the tool to call after previewing a bulk operation when you want to decline it rather than letting it expire. However, it does not explicitly mention alternatives or when-not-to-use conditions, so the guidance is not as strong as it could be.

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

• inbox_listA
  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 idempotentHint=true, so the safety profile is covered. The description adds valuable context by clarifying that results are scoped to the API key and that each entry includes UUID, email, display name, provider, optional brand, and capabilities—information that helps an agent select the correct inbox_id.

  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?

  Two sentences with no wasted words. The purpose is front-loaded, the critical 'call it FIRST' instruction is prominent, and the output payload is summarized efficiently without reproducing the schema.

  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?

  The description is complete for a simple read-only list tool with an output schema. It explains the tool's purpose, when to use it, what the response contains, and how the response relates to other tools. Nothing essential is missing 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.

  Parameters3/5

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

  Schema description coverage is 100%, and the schema already documents provider and include_capabilities well. The description does not add new param-level meaning beyond mentioning the capabilities object in the response, but it does not need to because the schema carries the load.

  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: 'List every inbox (mailbox or account) this API key may use.' It clearly distinguishes the tool's role from siblings by positioning it as the entry point that supplies the inbox_id for other tools. Each output field is listed, so an agent knows exactly what to expect.

  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 explicitly says to 'Call it FIRST for the inbox_id the other tools take,' giving clear contextual guidance on when this tool should be used. It does not discuss when not to use it or name alternatives, but no sibling serves the same discovery/list role, so this level of guidance is sufficient.

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

• approval_decideA
  4.4/5.0
  Behavior4/5

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

  Annotations already signal readOnlyHint=false and idempotentHint=true, so the description's main contribution is explaining the real-world effect: the send 'is never delivered'. It also discloses process constraints — only 'reject' is accepted and approval requires an owner or admin in a browser session. This is meaningful context beyond the structured annotations, though it stops short of stating whether the rejection can be undone.

  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 tight sentences with no filler. The primary action and consequence are front-loaded, and the second sentence covers the critical limitation (no approval via MCP) without digressing into schema details or sibling comparisons. Every clause 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 simple mutation with 3 parameters and no output schema, the description is largely sufficient: it clarifies what this tool does, what it cannot do, and how the only decision value behaves. It does not describe the success response or what happens after rejection, but the core decision an agent needs to make — when and how to call it — is fully covered.

  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 the input schema fully documents approval_id, decision, and note. The description adds useful emphasis that decision accepts only 'reject' and explains why approving is not available, but this largely mirrors the schema's own parameter-level description. Thus the description adds only marginal semantic value beyond the schema.

  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: 'Reject a send that is waiting for human approval', and adds the key consequence 'so it is never delivered'. It also distinguishes itself from approval actions by declaring that approving is not available over MCP. This makes the tool's unique role unambiguous against siblings like approval_review and approval_schedule.

  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 usage context: use this when a send is pending human approval and needs to be rejected. It also gives a clear when-not: approval is deliberately not offered here and requires a signed-in browser session at the review_url. This routes the agent away from misuse even without naming a sibling tool.

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

• approval_updateA
  4.4/5.0
  Behavior4/5

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

  Annotations already cover idempotency and lack of destructiveness; the description adds useful behavioral context by noting that the reviewer sees the edited version and that the message still requires approval afterward. This gives an agent a clearer picture of the edit's real-world effect beyond just a generic mutation.

  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, each earning its place: the first names the action, the second explains the post-edit workflow, and the third constrains the subject parameter. The most important information is front-loaded, and there is no filler.

  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 small update tool, the description plus full schema coverage gives an agent enough to invoke it correctly: what can be changed, the required approval_id, and the continuing approval requirement. A minor gap is that it does not describe the outcome if the approval is no longer pending, but this is not essential for a correct call.

  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 schema already describes all four parameters at 100% coverage, so the baseline applies. The description adds a meaningful nuance beyond the schema by explaining that 'subject' only applies to operations carrying one (email_send, schedule_create), which is directly relevant to when an agent should pass that parameter.

  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 ('Change') and resource ('a send that is waiting for human approval'), and clearly identifies the editable fields (subject or body). It also clarifies the scope of subject editing by naming which operations carry one, which helps distinguish this from approval_decide and approval_review.

  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 clearly indicates this is for pending sends still awaiting approval, and notes that approval must still happen afterwards, implicitly steering agents away from using this as a final decision tool. It does not explicitly list alternatives or say 'do not use for approved sends,' but the context is strong enough for correct routing among siblings.

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

• approval_reviewA
  4.5/5.0
  Behavior4/5

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

  Annotations already provide readOnlyHint, idempotentHint, and destructiveHint. The description adds useful behavioral context by explicitly stating it is read-only and by explaining that approval is not performed here but requires a separate review page. This goes beyond the annotations and clarifies a key operational boundary.

  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 sentences long, begins with the primary purpose, and every clause earns its place. It packs the key use case, the read-only nature, and the important limitation about approval without any filler.

  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 single-parameter, read-only tool with strong annotations, the description is complete. It tells the agent what will be retrieved (full contents including body), what not to expect (approval capability), and where to direct the user for approval. No output schema is present, but the description sufficiently clarifies the return intent.

  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 already covers the single parameter approval_id at 100%, including its type, format, and source context. The description does not add new parameter semantics beyond what the schema provides, so the baseline score 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 specific verb ('Fetch'), a specific resource (a send waiting for human approval), and the purpose (display in the review card). It clearly distinguishes itself from approval_decide by noting that approving is not possible here.

  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 use it: for a send pending approval, to show in the review card. It also gives a clear exclusion: approving is not possible from this tool and requires the signed-in review page linked as review_url. This helps an agent avoid choosing the wrong sibling.

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

• bulk_executeA
  4.6/5.0
  Behavior5/5

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

  Beyond the annotations (destructiveHint=true, idempotentHint=false), the description discloses crucial behaviors: the plan executes at most once, expires after 15 minutes, and the affected message set is immutable. This prevents an agent from retrying or modifying after the fact, which is not captured in annotations.

  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 sentences contain only high-value information with no repetition or filler. The core action is front-loaded, and critical constraints follow logically.

  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 single-parameter destructive tool with annotations covering the safety profile, the description is nearly complete. It might benefit from noting what a successful execution returns, but this is minor given the low complexity and no output schema.

  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 already documents plan_id with 100% coverage, but the description adds meaningful semantics: it is the only thing accepted, and the messages affected were frozen at preview time. This enriches the parameter beyond a simple schema declaration.

  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 ('Run') and resource ('bulk delete or move that was previewed as a plan'), clearly differentiating it from bulk_cancel or single-message deletes. The title and first sentence together make the tool's purpose unmistakable.

  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?

  Clearly indicates that execution follows a previewed plan and that the plan is frozen, so if an agent needs to change affected messages, this is not the tool. It does not explicitly name bulk_cancel or another alternative, but the constraint context strongly implies when to use it and the one-shot execution rule.

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

• contact_searchA
  4.7/5.0
  Behavior5/5

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

  Annotations already declare readOnly, idempotent, and openWorld, but the description adds critical non-obvious behavior: a bounded, header-only scan of a recent mail window, message_count counting only within that window (not all-time), and results ordered 'most recent first.' No contradiction with annotations.

  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?

  Four sentences, each earning its place: purpose, behavioral caveat, return format, and parameter guidance. The core purpose is front-loaded with no redundancy.

  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?

  With an output schema present and rich annotations, the description covers the remaining non-obvious semantics an agent needs—recent-window behavior, the meaning of message_count, and inbox_id omission—making the tool fully callable without further inference.

  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 description coverage is 100%, so the baseline is 3. The description adds meaningful nuance by explicitly explaining that omitting inbox_id scans every accessible inbox, reinforcing but extending the schema's 'set it only when the user named a specific inbox' guidance.

  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 and resource: 'Find people by name or email fragment.' It clearly distinguishes this tool from the email-centric siblings by describing contact search behavior and the 'no stored contact list' caveat, reinforcing what it is not.

  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?

  Provides clear context and a concrete usage rule: 'For general or cross-inbox questions ... OMIT inbox_id so every accessible inbox is scanned.' It does not name an alternative tool, but no contact-search sibling exists, so there is no alternative to exclude.

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

• automationA
  4.8/5.0
  Behavior5/5

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

  Annotations declare destructiveHint=true and readOnlyHint=false, so the description carries the burden of elaborating side effects — and it does so thoroughly. It discloses that rules are created DISABLED, delete keeps run history, preview is a DRY RUN, forward is ALWAYS held for approval, draft_reply only writes a draft, and label semantics vary by provider. This goes well beyond the annotation hints and prevents agent surprise.

  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?

  The description is dense but every sentence earns its place. It front-loads the core concept, then systematically covers actions, key naming, rule actions, safety constraints, and permissions. It is longer than average, but the tool's scope — nine operations with nested objects — justifies the length. No filler or tautology.

  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 tool with 10 parameters, nested objects, nine actions, and no output schema, the description is remarkably complete. The schema's conditional blocks handle per-action requirements, while the description supplies all interpretive context: filter criteria reference, raw-query rejection, interval ladder rationale, per-run blast radius, approval behavior, and permission requirements. Nothing an agent needs to invoke this safely is missing.

  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 baseline is 3. The description adds meaning well beyond the schema: it resolves the confusing action vs rule_action naming, explains preview's dry-run semantics, details label application differences (Gmail label, Outlook category, IMAP keyword), and documents draft_reply template substitution limits. The only minor deduction is that parameter semantics could have been even more explicitly mapped to each action.

  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+resource combination: 'Create and manage unattended scheduled triage rules.' It immediately clarifies the tool's nature — a stored search plus one fixed action, evaluated on a cadence with no model in the loop. It distinguishes itself from siblings like email_delete and email_compose by enumerating rule actions and explicitly stating deletion is unavailable.

  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 operational guidance: 'Always preview before you enable', notes that every action needs manage:automations, explains the two-key distinction between action and rule_action, and states when the tool is not appropriate (raw queries are rejected, deleting mail is refused). This is strong when-to-use and when-not-to-use guidance.

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