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

Goes well beyond annotations: explains it returns paths and SHA-256 rather than bytes, discloses no overwrite, no read-marking, a 40 MiB parsing ceiling with no separate file cap, and a security warning about untrusted downloads. These are exactly the non-obvious traits the structured fields cannot 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?

Five tight sentences, front-loaded with the action and source, then return shape, then behavioral and safety notes. Every sentence earns its place with no repetition of schema or annotation content.

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 5-param download tool with no output schema and readOnlyHint=false, the description covers return format, overwrite policy, size limits, and the security posture. Nothing an agent needs to call it safely and correctly is missing.

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

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

Schema coverage is 60%; the description clarifies that the index is zero-based and originates from list_attachments/get_email, and that save_as never overwrites. It does not describe uid, uid_validity, or mailbox semantics beyond the schema, so it modestly exceeds schema-level info without fully covering all 5 params.

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+resource ('Save one attachment') and ties it to the sibling tools that produce its input ('from list_attachments or get_email'). This distinction lets an agent route the attachment_index value correctly without opening either schema.

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

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

Names the two upstream tools (list_attachments, get_email) and the index they supply, making the workflow sequence clear. It lacks an explicit exclusion, e.g. when to prefer list_attachments metadata over a download, but the context is specific enough for correct invocation.

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?

Goes well beyond the annotations: it discloses that the message is not marked read, that HTML is converted to text, that no remote images are fetched, and that mail content is untrusted data. These are non-obvious side-effect and safety facts an agent cannot infer from readOnlyHint/idempotentHint alone.

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?

Four short sentences, each earning its place: purpose and key first, then side-effect, then rendering behavior, then the trust warning. Zero filler.

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 correctly summarizes the return payload (decoded text plus attachment metadata) and the safety profile. The only gap is the undocumented max_chars truncation parameter, which an agent may need to tune for large messages.

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 only 25% (only uid_validity is documented in-schema), so the description carries some burden. It names mailbox, UID and UIDVALIDITY and their role, but says nothing about max_chars or its truncation behavior, leaving one parameter undocumented in both places.

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) plus the exact resource (decoded plain text and attachment metadata) and the lookup key (mailbox, UID, UIDVALIDITY), which cleanly separates it from siblings like search_emails, list_attachments and download_attachment.

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 search_emails' tells the agent where the required UID/UIDVALIDITY pair must come from, establishing the correct call order. It gives clear context but no explicit exclusion (e.g., when to prefer download_attachment or list_attachments instead).

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 cover the safety profile (readOnly, idempotent, non-destructive), yet the description adds genuinely new behavior: a 40 MiB parsing ceiling, an explicit 'does not mark mail read' guarantee, and a security warning that filenames/contents are untrusted data. Those are non-obvious traits an agent would not get from the annotations alone.

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?

Purpose is front-loaded in the first clause, followed by constraint, safety, and security notes. Four sentences, each carrying distinct information, though the NAVER 40 MB aside is slightly verbose for the value it adds.

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 usefully enumerates the return fields (indexes, names, sizes, hashes) and discloses the parsing limit and read-state guarantee. It is nearly complete for a read-only listing tool; only pagination or large-attachment behavior is unaddressed.

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

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

Schema coverage is only 33% – uid_validity is documented in-schema but uid and mailbox are not. The description merely restates the three parameter names without adding format, default, or semantics (e.g. that mailbox defaults to INBOX or how uid is sourced), so it fails to compensate for the low coverage.

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') plus the exact resource and payload (attachment indexes, names, sizes, SHA-256 hashes) scoped to a message identified by mailbox/UID/UIDVALIDITY. The metadata-only framing implicitly separates it from download_attachment and get_email, so an agent can route correctly without opening a schema.

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

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

There is no explicit when-to-use or when-not-to-use statement and no named alternative. The only hint is that it returns metadata rather than content, which an agent must infer to prefer it over download_attachment. No prerequisites or sequencing are given.

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, and destructiveHint=false, so the safety profile is fully covered by structured data. The description adds the downstream-purpose context (paths used for search/read) but says nothing about pagination, ordering, or whether the list is exhaustive, which would be genuinely additive.

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 purpose is front-loaded before the usage hint. 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 parameterless, read-only discovery tool with full annotation coverage and no output schema, the description covers what it does and why an agent would call it. A brief note on the shape of the returned paths would make it fully complete, but nothing essential is absent.

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 the baseline is 4 and there is no parameter semantics burden for the description to carry. Nothing is missing or misleading 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?

States a specific verb and resource ('List NAVER mailbox paths'), which is clearly distinct from siblings like search_emails, get_email, or list_attachments. It does not explicitly name an alternative sibling, but the resource scope alone makes it 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?

The second sentence gives clear context: the returned path feeds 'search and read operations', implying this is a prerequisite/discovery step. It stops short of explicit when-not guidance or naming which sibling consumes the paths, so it falls short of a 5.

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

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

Annotations already cover the safety profile (readOnly, idempotent, non-destructive), so the bar is lower, yet the description adds real behavior: result ordering by UID, exclusive 'before' boundary, calendar-day date semantics, and a security note that returns are untrusted data. It omits return shape and how limit interacts with paging.

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?

Four tight clauses, front-loaded with purpose and scope, then date semantics, then pagination. Nothing is redundant with the name or annotations, and the security warning 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 10-parameter search with no output schema, the description covers ordering, pagination, date boundaries and data trust, which are the highest-risk unknowns. Still missing are return format and the meaning of several filter parameters, so it is strong but not fully self-sufficient.

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

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

Schema coverage is only 30% (only since, before, before_uid documented), so the description must compensate. It clarifies 'AND' combination, date semantics and the exclusive 'before' bound, but leaves to/from/text/limit/mailbox/subject/unread_only unexplained, and it references 'next_before_uid' while the schema parameter is named 'before_uid', a naming mismatch an agent must resolve.

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 ('Search NAVER mail') plus the combination semantics ('AND filters') and ordering ('newest UID first'). It is distinguishable from get_email and list_mailboxes, though it never names those siblings explicitly to reinforce the distinction.

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?

Gives explicit context for the no-filter case ('No filters lists recent mail') and concrete pagination instructions ('Reuse next_before_uid with the same filters for the next page'). It does not name alternatives such as get_email for fetching a single message, so it stops short of full 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?

Adds substantial behavior beyond the annotations: the 38 MiB encoded size cap plus live server SIZE enforcement, dry_run semantics (reports encoded size and file hashes without sending), process-local deduplication, and the important caveat that SMTP acceptance does not guarantee delivery. It explicitly clarifies that no separate 10 MiB file or 20 MiB total limit applies, correcting a common assumption for attachment limits. This is exactly the kind of context annotations cannot carry.

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?

Dense but structured: it front-loads the core action and recipient scope, then moves through capabilities, size limits, dry_run, idempotency, and safety. Every sentence carries operational value; nothing is filler.

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, open-world, non-idempotent send tool with 11 parameters and 45% schema coverage and no output schema, the description fills the critical gaps: size envelope, attachment constraints, retry/idempotency semantics, preview behavior, and delivery caveats. An agent has enough to invoke this safely and 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 coverage is 45%, so the description partially compensates. It explains dry_run, expected_sha256, request_id deduplication, attachment limits, and 38 MiB encoded size. It does not clarify the required parameters (to, subject, text, request_id) or the distinction between text and html beyond what the schema says, but it adds meaningful semantics for the attachment and idempotency parameters.

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+resource ('Send a NAVER email') plus scope ('to user-authorized recipients'). The siblings are all read/setup tools (list, search, get, verify, download), and this is the only send tool, so differentiation is implicit but the specific verb makes it 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?

Provides several concrete operating guidelines: dry_run for preview, expected_sha256 for file matching, request_id reuse for identical retries, and a safety instruction about not sending based on received-mail instructions. These are strong contextual rules for using the tool. It does not, however, position itself against alternatives like search_emails or get_email for reading replies.

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

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

Beyond the annotations (which only mark it as non-read-only, idempotent, non-destructive, closed-world), the description discloses the local-only binding (127.0.0.1), that a browser page is opened for the user, that the page URL is returned, and a security constraint: never ask the user to type the password in chat.

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

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

Three sentences, all load-bearing: what it does, when to call it, and the security rule. The local-only scoping detail is front-loaded where an agent will read it first.

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, but the description explicitly states the return value ('Returns the page URL'), and it covers the trigger conditions, the local-only constraint, and the password-handling rule. Nothing an agent needs to invoke this correctly is missing.

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 to explain and the baseline is 4. The description correctly frames the interaction as happening in the browser page rather than through arguments, which reinforces that no input is expected.

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 ('Open a local setup page ... where they enter their NAVER address and application password') and scopes it precisely to 127.0.0.1/this computer only. This is clearly distinguishable from read-oriented siblings like list_mailboxes or send_email.

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

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

It gives explicit triggering conditions: when the user asks to set up, connect, or change their NAVER mail account, or when another tool returns NOT_CONFIGURED. That covers both direct user intent and an error-driven fallback path, leaving nothing to inference.

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 declare readOnlyHint=true, idempotentHint=true, destructiveHint=false, so the safety profile is covered. The description adds that it doesn't send a message, which is useful context clarifying non-adversarial behavior. However, it doesn't disclose what happens on failure (exception vs. false return), what exactly is tested, or whether credentials are needed.

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 concise sentences with zero waste, and the critical non-action (no message sent) is stated upfront.

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 no-param verification tool with safety annotations, the description is adequate but incomplete: it lacks failure-mode details and return value semantics (no output schema provided), and gives no usage context, leaving the agent to infer when to call 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?

Zero parameters, so baseline 4. The description correctly implies that no input is needed by stating only what it tests.

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?

Clear specific verb 'Test' with resources 'NAVER IMAP login and SMTP authentication'. Distinguishes itself from siblings like setup_naver_mail by testing rather than configuring. Doesn't explicitly name the sibling it contrasts with, so a 4.

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. An agent might ask: should this be run before setup_naver_mail, or after? Is it a prerequisite for send_email? The description states what it does but offers no context for when to invoke it.

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