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

Annotations already declare readOnlyHint=true and openWorldHint=true, so the safety profile is covered; the description's note that search=true attaches live web results usefully reinforces the open-world nature. It adds nothing about latency, whether conversation state is persisted server-side, or what the call returns, which would be genuinely new behavioral information.

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 sentences, front-loaded with the core action, then the two scoping controls. Every sentence carries distinct information and nothing is repeated from 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?

For a 7-parameter tool with no output schema, the description leaves a key chaining gap: it tells the agent to pass parentMessageId but never indicates where that id comes from (presumably the response). It also says nothing about refFileIds pairing with upload_files or when the vision model applies, though the schema covers those individually.

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%, so every parameter is already documented, and the baseline is 3. The description does add modest relational meaning by pairing sessionId with parentMessageId to continue a thread, but leaves model, thinking, and refFileIds untouched beyond what the schema says.

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 ('Send a prompt to DeepSeek'), making it immediately clear this is the chat-completion tool. It is distinguishable from siblings like new_session or web_search through the explicit note about the search flag, though it never names those siblings directly.

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 two concrete use conditions: set search=true for live web results, and pass sessionId + parentMessageId to continue a conversation. That is clear when-to guidance, but there are no explicit exclusions or named alternatives (e.g., use web_search alone vs. chat with search), 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 declare readOnlyHint=true and openWorldHint=true, so safety and the fact that it reaches out to a remote service are covered. The description adds one useful nuance — it checks both configuration and acceptance, not just presence — but says nothing about latency, failure modes, or what an unconfigured/unaccepted result looks like.

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, and the operative detail ('configured and accepted') lands immediately. Nothing could be trimmed 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 zero-parameter, read-only health check with no output schema, the description conveys enough to call it and interpret the answer as a pass/fail on credential validity. It could go slightly further by describing the shape of a negative result, 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 there is nothing for the description to disambiguate. The baseline for a parameterless tool applies; no parameter information is missing because none exists.

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 ('Report') and a specific subject ('whether the DeepSeek credential is configured and accepted'), which is clearly distinct from file, chat, and session siblings. It stops short of naming the closest sibling (refresh_token), so the boundary between 'diagnose' and 'fix the credential' must be inferred.

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: an agent can guess this is a pre-flight or troubleshooting check before calling chat, but the description never says when to call it or when to prefer refresh_token instead. No exclusions or prerequisites are stated.

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 and openWorldHint=true, so safety is already covered. The description adds useful behavioral context by listing exactly what metadata is returned, including upload status and content-audit verdict, which helps the agent understand the response shape.

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 compact sentences, front-loaded with the action and resource, and the second sentence provides usage context without repetition or 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 simple read-only lookup with one fully documented parameter and annotation-covered safety, the description is complete. It also compensates for the absence of an output schema by naming the returned metadata fields.

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 100% and the single fileIds parameter is already documented as coming from upload_files. The description reinforces that these are previously uploaded file ids, but adds little meaning beyond the schema.

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: fetch server-side metadata for previously uploaded file ids. It also names the exact metadata fields returned, which clearly distinguishes it from the upload_files sibling.

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?

Explicitly says to use it to confirm an upload finished or re-check status before attaching. This gives clear use context, though it does not name explicit exclusions or alternative tools beyond the implied upload_files relationship.

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 and openWorldHint=true; the description usefully discloses that the session starts empty and that the response carries an id and URL, which matters since there is no output schema. There is mild tension between 'Create' and readOnlyHint=true, but for an ephemeral, non-destructive session this reads as a harmless side effect rather than a contradiction; no auth, rate-limit, or lifecycle details are given.

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 action and the return payload are both stated immediately.

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, no-output-schema tool, the description covers what it creates and what comes back (id and URL), which is the essential information. It stops short of saying what the URL is for or what the agent should do with the id next.

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 compensate for; baseline 4 applies. The description correctly signals that no configuration is required to open the session.

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 ('Create an empty DeepSeek chat session') plus what is returned. It is clearly distinct from siblings like chat, upload_files, and list_files, though it never explicitly says how it relates to the chat tool it presumably feeds.

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: an agent can infer you create a session before chatting, but the description never states when to call this versus 'chat' or whether an existing session can be reused. No prerequisites, no alternatives, no exclusions.

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?

The description says the tool saves the token ('save it'), a state-mutating action, while annotations declare readOnlyHint=true. That is a direct conflict about side effects; the description also says nothing about behavior when no signed-in browser profile exists.

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, action first and the trigger condition second, with no filler or repetition of the title.

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?

The description covers what the tool does and when to call it, which is most of what an agent needs. However, with no output schema, it should state what a successful refresh returns and what happens when the credential capture fails, and it never reconciles the persistence side effect with the readOnly annotation.

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 no parameters, so there is nothing for the description to explain; the baseline for a zero-parameter tool 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: re-capture the DeepSeek bearer token from the signed-in browser profile and persist it. That is distinct from every sibling (diagnose, chat, list_files), none of which touch credentials.

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 an explicit trigger: 'Use when another tool reports an auth rejection.' It does not say when NOT to use it or whether diagnose should be run first to confirm the failure is auth-related, so routing is clear but not fully disambiguated.

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=false, destructiveHint=false and openWorldHint=true, so the write-scope is covered. The description adds genuinely new behavior: it persists artifacts into the account, returns a file id per file, and enforces a 104857600-byte per-file cap — all things the annotations cannot convey. It omits only partial-failure behavior when some paths fail.

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 sentences, zero filler, front-loaded with the action and resource. The downstream refFileIds instruction and the size limit both earn their place as decision-relevant facts.

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 one-parameter mutation with annotations and no output schema, the description supplies the missing pieces: return value (file id), consumption path (refFileIds), and hard limit. Only the multi-file atomicity question — whether one bad path fails the whole batch — is left 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?

With one parameter at 100% schema coverage the baseline is 3, and the schema already documents 'Absolute or relative local paths'. The description earns a bump by constraining what those paths may point at (images, PDFs, documents, code, archives) and by stating the per-file size ceiling, which directly tells the agent how to split or reject inputs.

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 ('Upload') and resource ('local files ... to the DeepSeek account'), enumerates accepted file categories, and is clearly distinct from the sibling list_files and chat tools. An agent can identify this as the write/ingest path 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.

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

Explicitly names the downstream step: 'pass those ids as refFileIds to the chat tool to attach them to a prompt,' which routes the agent correctly between upload and chat. It gives the positive workflow but stops short of stating exclusions (e.g., don't re-upload an existing file id, or what to do for remote URLs).

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 and openWorldHint, so the safety/external-call profile is covered. The description adds value beyond that by disclosing the return shape (answer plus session link), which matters because there is no output schema. It does not mention latency, cost, or how sessions are created/managed.

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 sentences, front-loaded with the core action, then usage triggers, then return value. No filler or redundancy.

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 two-parameter, read-only search tool with no output schema, the description covers purpose, triggers, and return content. Minor gaps remain around session handling and the effect of the `thinking` flag, but nothing prevents correct invocation.

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%, so both `query` and `thinking` are already documented, and the description adds no extra meaning about the reasoning toggle or query formatting. Baseline 3 applies when the schema does the heavy lifting.

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 action and mechanism: ask DeepSeek with web search enabled, returning an answer plus a session link. It implicitly separates itself from the plain `chat` sibling by emphasizing live-web grounding, though it never names that sibling 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?

Gives concrete use cases (current events, prices, releases, anything requiring up-to-date sources), which is a clear trigger condition. It stops short of stating when NOT to use it or pointing at `chat` for non-web questions, so there is no explicit alternative routing.

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