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

Annotations already declare the write/safety profile (readOnlyHint=false, destructiveHint=true, idempotentHint=false, openWorldHint=false). The description adds genuinely useful context -- drafts persist across MCP restarts and are distinct from browser drafts -- but it omits why destructiveHint is set (e.g. overwrite behavior) or any auth/account requirements.

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

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

Two tight sentences with the core action front-loaded and no filler. It is well-sized, though the second sentence could be trimmed slightly without losing meaning.

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

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

For a 6-parameter mutation tool with no output schema and zero parameter descriptions anywhere, the definition leaves critical gaps: required fields (to, subject), body content selection (html vs text), attachment handling, and the non-idempotent/destructive behavior are all unaddressed. Persistence is covered, but an agent lacks the information to invoke it correctly.

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

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

Schema description coverage is 0% across 6 parameters, including a nested attachments object with base64 content, and the description says nothing about any of them (to, subject, html/text body, accountId, attachments). With low coverage the description is expected to compensate, and it does not.

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

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

States a specific verb and resource ('Save a local draft'), and the qualifier 'local' plus 'separate from browser drafts' clarifies the domain. It does not differentiate from the sibling write tools (update_draft, delete_draft), 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.

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

There is no guidance on when to use create_draft versus update_draft, or whether a draft should be created before sending. The persistence and browser-draft statements are contextual facts, not usage routing, so an agent gets no when-to-use/when-not signal.

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

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

Annotations already declare destructiveHint=true, idempotentHint=true, and openWorldHint=false, so the safety profile is largely covered structurally. The phrase 'using its current revision' adds genuine behavioral context by implying optimistic-concurrency semantics, but the description never says what happens on a revision mismatch, whether deletion is recoverable, or that the operation is irreversible — gaps the agent would want filled for a destructive tool.

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

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

One short sentence with no filler, and the core action leads. It is arguably under-specified rather than verbose, but as a structural matter it is tight and front-loaded.

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

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

For a destructive mutation with 0% parameter coverage, no output schema, and two required params, the description omits concurrency failure behavior, recoverability, and any post-delete state. The annotations supply the safety hints, but the description does not complete the picture an agent needs to invoke this safely.

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

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

Schema description coverage is 0%, so the description carries the full burden, and it only gestures at the 'revision' parameter ('its current revision') without specifying that it must match the draft's current revision or what a mismatch produces. The draftId parameter gets no explanation at all beyond the name, leaving both required params thinly documented.

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

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

The description states a specific verb and resource, 'Delete a local draft', which is unambiguous against siblings like create_draft, get_draft, and update_draft. It adds the qualifier 'local draft' and the revision condition, but does not explicitly contrast itself with any sibling, so it falls short of a 5.

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

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

There is no when-to-use guidance, no statement of prerequisites such as needing the latest revision fetched via get_draft, and no mention of alternatives. The only implied usage is that the draft must be local and have a revision.

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

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

Annotations already declare readOnly=true, idempotent=true, destructive=false and openWorld=true, so safety is covered. The description adds genuinely new behavior: the payload comes back as Base64 and is bounded by MAIL_MAX_ATTACHMENT_BYTES, a size constraint the annotations do not express.

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

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

One front-loaded sentence carrying the action, encoding, and limit with zero filler. Every clause earns its place.

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

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

For a simple read tool with no output schema and full annotation coverage, the description adequately covers format and size cap. It still leaves open what happens when the attachment exceeds MAIL_MAX_ATTACHMENT_BYTES (failure vs truncation) and leaves all four parameters unexplained.

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

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

Schema description coverage is 0% across four parameters (emailId, attachmentId, folder, accountId), and the description compensates only faintly by implying an attachment belongs to a message it owns. It says nothing about the inbox/sent enum, the accountId scoping, or required-vs-optional behavior.

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

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

The description states a specific verb and resource ('Read an attachment from an owned message') plus the return encoding (Base64), which is enough to distinguish it from list_attachments. It does not name or contrast any sibling explicitly, so it falls short of the 5 bar.

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

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

'from an owned message' hints at a scope prerequisite, but there is no statement of when to reach for this tool versus list_attachments (discover the attachmentId first) or get_message. No alternatives or exclusions are offered.

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

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

Annotations already declare readOnly, idempotent, non-destructive and closed-world, so the safety profile is covered. The description adds that content and attachment data are included in the result, which is useful, but it says nothing about behavior for an unknown or deleted draftId, or whether attachments are inlined versus referenced.

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

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

One front-loaded sentence with no filler; the return payload is stated before any hypothetical detail. It is arguably too terse for the surrounding draft CRUD family, but nothing in it is wasted.

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

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

For a single-parameter read with no output schema, the description is minimally sufficient: it says what comes back, which is the main thing an agent cannot infer. It stops short of covering the not-found case and the local-versus-remote distinction that the sibling set makes relevant.

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

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

Schema description coverage is 0%, so the sole parameter draftId carries only a uuid format and pattern in the schema. The description never explains that draftId is the local draft identifier or where to obtain it, but with a single self-evident ID parameter this is a minor gap rather than a blocking one.

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

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

States a specific verb ('Read') and resource ('a local draft') plus the payload the caller gets back ('content and attachment data'). The word 'local' distinguishes it from a server-side message fetch, though it never names list_drafts or get_message explicitly.

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

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

Usage is only implied: 'local draft' plus a required draftId suggests fetching one known draft rather than enumerating, which is what list_drafts is for. There is no explicit when-to-use statement, no exclusions, and no mention of how this differs from get_message or create_draft/update_draft.

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

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

Annotations already declare readOnlyHint, idempotentHint, destructiveHint=false and openWorldHint=true, so the safety profile is covered. The description adds genuinely useful non-obvious behavior: it returns text and HTML content, and critically that reading does not mark the message read — an effect annotation hints at but does not state.

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

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

Three short, front-loaded sentences with no filler: purpose first, then the important side-effect disclosure, then the folder default. Every sentence earns its place.

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

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

For a read-only tool with no output schema and annotations covering safety, the description is adequate on behavior and content. The gap is parameter coverage: with 0% schema descriptions, emailId and accountId are undocumented anywhere, which matters for a tool needing to locate a specific message.

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

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

With 0% schema description coverage and 3 parameters, the description carries the burden and does not meet it. The only parameter detail given ('folder defaults to inbox') merely repeats the schema's default value, and neither emailId nor accountId is explained at all, so an agent gets no guidance on identifying the message or selecting an account.

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

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

States a specific verb ('Read') and a precisely scoped resource ('one owned message including text and HTML'), which distinguishes it from list_messages and search_messages by signaling single-item retrieval. It does not name a sibling explicitly, so the differentiation is implied rather than stated.

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

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

'Read one owned message' implies this is the single-item lookup versus the list/search siblings, and the folder default hints at mailbox context. However, there is no explicit when-to-use or when-not-to-use guidance and no named alternative for scenarios like retrieving many messages.

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

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

Annotations already declare readOnly, idempotent, non-destructive, closed-world behavior, so the bar is lower. The description adds genuinely useful context beyond them: the receipt is local (not an authoritative server round-trip) and defines two return states, 'accepted' as API acceptance and 'pending' as requiring reconciliation.

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

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

Two dense sentences, zero filler, with the core action and the meaning of each status front-loaded in the order an agent needs them.

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

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

No output schema exists, so the description should convey the return states; it explains 'accepted' and 'pending' but does not enumerate other possible states (e.g., rejected, unknown) or the staleness implications of reading a local receipt. For a one-parameter read tool this is close to complete.

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

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

Schema description coverage is 100% and the operationKey description already covers stability, reuse on retry, and immutability after uncertain results. The description confirms the key scopes a single send but adds no syntax or format detail beyond the schema, so the baseline 3 applies.

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

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

States a specific verb and resource ('Read the local receipt for an operationKey') and the word 'local' implicitly contrasts with server/API status checks. It does not, however, name or differentiate itself from the sibling mail_status, which an agent could easily confuse with a send-status lookup.

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

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

The description gives an operational rule tied to a result state ('pending requires reconciliation before any retry'), which implies this is the tool to call after an uncertain send. It never states when to prefer it over mail_status or what to do if the key is unknown, so selection guidance is only implied.

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

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

Annotations already declare readOnlyHint=true, idempotentHint=true, destructiveHint=false and openWorldHint=true, so the safety profile is fully covered. The description adds only that results are 'metadata' (not content) and that the message must be 'owned' — modest but non-redundant context; nothing is said about pagination, empty results, or ordering.

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

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

A single efficient sentence with the resource front-loaded and zero filler. It is appropriately sized for a one-line tool, though its brevity contributes to the gaps noted elsewhere.

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

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

With no output schema, no annotations gap on safety, and three parameters at 0% schema coverage, the description should describe what attachment metadata is returned (name, size, mime type, id) and how folder/accountId affect the lookup. It supplies none of this, leaving the agent unable to predict the response.

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

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

Schema description coverage is 0% and the description names none of the three parameters. The inbox/sent 'folder' enum and the 'accountId' scoping parameter are completely undocumented anywhere, and 'emailId' is only hinted at via 'owned message'. The description fails to compensate for the coverage gap.

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

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

States a specific verb and resource ('List attachment metadata') and scopes it to an owned message, which implicitly contrasts with download_attachment's content retrieval. However, it never names the sibling or clarifies the metadata-vs-content boundary explicitly, so an agent must infer the differentiation.

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

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

No when-to-use guidance, no prerequisites (e.g. that an emailId from list_messages/get_message is required), and no mention of download_attachment as the alternative for retrieving actual file contents. The agent is left to infer the call context entirely.

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

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

Annotations already declare readOnlyHint, idempotentHint, non-destructive and closed-world, so the safety profile is fully covered. The description adds one genuinely useful behavioral fact — that the listing is limited to locally stored drafts — but says nothing about ordering, limits, or result size.

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

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

A single front-loaded sentence with no filler; the scope qualifier arrives immediately after the verb and resource. Nothing in it is redundant with the title or schema.

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

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

For a zero-parameter read tool with full annotation coverage this is close to sufficient, but there is no output schema, so the description is the only place an agent could learn what a 'draft' entry contains, in what order results come back, or whether the list is bounded.

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

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

The tool takes zero parameters, so there is no parameter semantics for the description to carry; the baseline for a no-param tool is 4. The description adds no misleading input expectations.

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

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

States a specific verb and resource ('List drafts') plus a scope qualifier ('stored locally for the authenticated user'), which separates it from server-side listing tools. It does not name the sibling it contrasts with (e.g., list_messages or get_draft), so an agent must infer the boundary from the resource noun alone.

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

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

Usage is only implied by the name and the 'local drafts' scoping — the description never says when to call this versus get_draft for a single draft or list_messages for the inbox. No prerequisites, no exclusions, no alternatives are named.

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

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

Annotations already declare readOnlyHint, idempotentHint, destructiveHint=false and openWorldHint, so the safety profile is covered. The description adds useful pagination context ('pass nextCursor fields for the next page'), which annotations do not convey. However, it refers to 'nextCursor fields,' a name that appears nowhere in the schema, which muddies rather than clarifies the actual cursor fields.

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

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

Two tight sentences with the core action front-loaded and pagination second. No filler, though the second sentence is spent on a field name that does not exist.

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

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

No output schema exists, so the description could have described the returned shape and how paging terminates, but it does not. With all three parameters undocumented and an incorrect cursor field reference, an agent lacks what it needs to page correctly.

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

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

Schema description coverage is 0% and none of the three parameters (limit, lastSort, afterAccountId) are explained: no meaning, ranges, or defaults. The only nod to parameters is 'nextCursor fields,' which does not match any actual parameter name, so it does not compensate for the coverage gap and may actively mislead.

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

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

Names a specific verb and resource ('List owned mailbox addresses'), so an agent immediately knows this returns mailbox addresses rather than messages or drafts. It does not explicitly name a sibling or state how it differs from them, but the resource is distinctive enough within the sibling set to be unambiguous.

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

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

There is no guidance on when this tool should be chosen, no prerequisites, and no mention of alternatives among the many list_* siblings. The single pagination sentence is procedural, not routing guidance.

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

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

Annotations already declare readOnlyHint, idempotentHint, destructiveHint=false, so the safety profile is covered. The description still adds two non-obvious behavioral facts the annotations do not: results are ordered newest first, and reading does not mark messages as read. It omits pagination/return-shape behavior, so not a 5.

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

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

Two short sentences, zero filler, and the scope/ordering constraint is front-loaded before the side-effect note.

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

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

For a simple list tool with all-optional parameters and no output schema, the description covers scope, ordering and read-state side effects, and annotations cover safety. It is still incomplete on paging (beforeId) and folder/account semantics, which an agent needs to call it correctly.

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

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

Schema description coverage is 0% and there are four parameters (limit, folder, beforeId, accountId), so the description carries the full burden. It says nothing about folder selection, result limits, beforeId-based paging, or account scoping, leaving every parameter semantically bare.

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

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

States a specific verb (list) and resource (messages in one mailbox folder) plus ordering, which distinguishes it from list_mailboxes and get_message. It does not explicitly differentiate from search_messages, so it falls short of a 5.

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

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

Usage is implied by the scope phrase 'one mailbox folder, newest first', which suggests a browsing/fetch-recent flow rather than a query. There is no explicit when-to-use, when-not-to-use, or named alternative such as search_messages.

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

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

Annotations already declare readOnlyHint, idempotentHint, openWorldHint, and destructiveHint=false, so the safety profile is covered. The description adds some value by scoping what is reported (identity, permissions, MCP configuration), but says nothing about auth requirements, cost, or failure modes for an open-world call.

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

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

A single front-loaded sentence with no filler, restatement of the name, or trailing caveats. It is exactly the right size for a zero-parameter status tool.

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

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

With no output schema, the description must convey what comes back; it does list the three categories of returned information, though only at a high level. For a parameterless read-only status call this is nearly complete, missing only a sense of when an agent should reach for it.

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

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

The tool takes zero parameters, so there is nothing for the description to disambiguate beyond what the schema establishes. Baseline 4 applies for a parameterless tool; the description neither helps nor hurts here.

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

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

The description names a specific verb ('Get') and a concrete set of resources: authenticated mailbox identity, permissions, and MCP configuration. It is unambiguous about what is returned, though it does not contrast itself with siblings like list_mailboxes or list_messages.

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

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

There is no guidance on when to call this tool versus the many sibling list/get tools, and no mention of prerequisites or typical contexts (e.g., verifying auth before other operations). Usage must be inferred entirely from the tool name and returned subject matter.

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

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

With readOnlyHint/idempotentHint already declaring the safety profile, the description still earns credit by disclosing pagination behavior (partial results signaled by exhausted=false, resume via nextBeforeId) and a real caveat that body search returns raw HTML. The cursor field is called 'nextBeforeId' while the schema parameter is 'beforeId', which introduces some naming ambiguity.

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

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

Three dense sentences with zero filler; the core matching rule and the partial-result rule are front-loaded. Nothing is redundant with the title or the schema.

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

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

There is no output schema, so the description must carry return-value semantics, and it does explain the exhausted flag and the resume cursor but not the result shape or ordering. For a 7-parameter, open-world search tool, the missing field/folder/limit semantics leave it only minimally complete.

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

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

Schema description coverage is 0% across 7 parameters, so the description carries the burden but only covers the query semantics (case-insensitive substring) and body matching. Field, folder, maxPages, maxResults, beforeId, and accountId are left entirely unexplained, including their defaults and bounds, for a search tool where field/folder selection is the main scoping decision.

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

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

States a specific verb (search), resource (owned mailbox pages), and matching mechanism (case-insensitive substring), so the operation is immediately legible. It does not, however, differentiate itself from the sibling list_messages or get_message, so an agent must infer which retrieval tool fits which case.

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

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

There is no statement of when to use search_messages versus list_messages or get_message, and no prerequisites (e.g. whether accountId is required when multiple accounts exist). The continuation hint ('exhausted=false ... continue with nextBeforeId') is behavioral rather than a use-case guideline.

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

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

Annotations already declare destructiveHint=true, readOnlyHint=false, and non-idempotency, so the safety profile is covered. The description adds genuine extra context by explaining that the revision is used for optimistic concurrency ('to prevent lost edits'), but it omits the failure mode when the revision is stale and the fact that the entire draft is overwritten.

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

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

A single front-loaded sentence with no wasted words, and the concurrency rationale is embedded efficiently rather than padded out.

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

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

For a destructive mutation with three required parameters, a nested message object, and no output schema, the description is too thin. It does not say what happens on a revision conflict, whether the replacement is total, or how the nested message payload relates to the existing draft.

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

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

Schema description coverage is 0% for the top-level parameters, so the description carries the burden. It clarifies the purpose of 'revision' but says nothing about 'draftId' or the complex nested 'message' object (to/subject/html/attachments), leaving most parameters unexplained.

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

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

States a specific verb and resource ('Replace a local draft'), which distinguishes it from the create_draft/delete_draft siblings. However, it does not name an alternative directly, and 'Replace' versus the tool name 'update_draft' creates mild ambiguity about full-replacement vs partial-update semantics.

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

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

There is no when-to-use guidance or routing to siblings such as create_draft or get_draft. The clause about 'current revision' hints at a concurrency workflow but does not tell the agent when to prefer update over create or what to do first.

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