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

Annotations already establish readOnly, idempotent, openWorld, and non-destructive behavior, so the bar is lower; the description still adds meaningful context by disclosing that the result includes credential provenance and remaining Admin API quota, hinting at a live account-level call. It stops short of saying whether the call itself consumes quota or requires specific credentials.

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 sentence, front-loaded with the resource, and no filler. The parenthetical list of returned fields is dense but every item is informative; the only nit is the clipped, verb-less phrasing.

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 zero parameters, full annotation coverage, and a dedicated output schema, the description needn't explain return structure and wisely doesn't. The remaining gap is the absence of any routing guidance against the closely named cloudinary_info sibling and no note on quota cost or auth requirements.

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 there is nothing for the description to disambiguate; baseline is 4. The description correctly adds no parameter chatter and instead spends its words on returned content.

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 enumerates exactly what the tool returns: cloud name, folder mode, credential source, remaining Admin API requests, and safety settings. That makes the resource specific enough to distinguish it from most siblings, but it never states a verb or explicitly contrasts with cloudinary_info, the one sibling an agent could plausibly confuse it with.

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?

Use is implied rather than stated: an agent can infer this is the place to check account identity, credential origin, or quota headroom, but the description never says when to call it, when not to, or why to prefer it over cloudinary_info. No alternatives or prerequisites 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 declare destructiveHint=true and idempotentHint=true, but nothing in structured data communicates the two-phase safety protocol. The description explains that the first call only returns needs_confirmation with a plan and confirm_token, and that nothing is deleted until the call is repeated with that token — the single most important behavioral fact for an agent invoking this.

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 sentences, no filler. The destructive scope comes first and the confirmation safeguard second, which is the correct reading order for a delete 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?

Six parameters with full schema coverage, an output schema for return shape, and annotations for the safety profile. The description fills the one remaining gap — the confirmation workflow — so an agent knows exactly what happens on first and second call.

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%, so the baseline is 3, but the description adds meaning the schema does not: recursive is framed as 'a folder and everything in it', and confirm_token is tied to the confirmation round-trip ('runs exactly that plan'). dry_run and the filter params are left to the schema, which documents them adequately.

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 (delete) and enumerates the three target kinds it covers: a single asset, wildcard matches, or a folder tree with recursive. No sibling tool does deletion, so the resource+verb pairing is unambiguous against cloudinary_move, cloudinary_rename_folder, and the read tools.

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 recursive case is called out inline as the condition for folder-tree deletion, which gives real usage context. It does not name a sibling or state when NOT to use it (e.g., prefer cloudinary_move to relocate), but for a destructive delete the use case is self-evident.

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 real behavior beyond the annotations: existing files are skipped by default, overwrite must be explicitly true, and writes are confined to allowed local directories. Annotations already flag destructiveHint/openWorld/idempotent, so the description's job is context, and it delivers the key mutation semantics, though it does not say what a wildcard match produces or whether failures are reported per asset.

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 sentences, zero waste; the skip/overwrite caveat is attached directly to the action sentence and the identifier scope is 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?

With a full schema, an output schema, and rich annotations, an agent has what it needs to call this safely. The remaining gap is wildcard semantics (single vs multiple resulting files) which the description leaves unstated.

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 the schema already documents all six parameters including enums and defaults; baseline 3 applies. The description only restates public_id forms and adds the 'allowed local directories' restriction on destination.

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 (Download) and resource (assets), and enumerates accepted identifier forms (name, public ID, wildcard) plus the destination constraint of allowed local directories. It is clearly distinguishable from cloudinary_upload and cloudinary_get_url by the local-file outcome, though no sibling is named.

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?

Implied usage only: it scopes placement to allowed local directories and notes skip-vs-overwrite behavior, which hints at bulk/local-mirror scenarios. It never says when to prefer this over cloudinary_get_url (URL only) or how it relates to cloudinary_list/tree for discovering public IDs.

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, and destructiveHint=false, so the safety profile is covered. The description adds the useful behavioral nuance that the transformation is optional and shows representative transformation strings, though it does not discuss signing/auth implications that the schema hints at.

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 sentence, front-loaded with the core purpose and 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?

An output schema exists so return values need not be explained, and all five parameters are documented in the schema. The description is complete enough for a simple read-only URL builder, with only minor gaps like signing behavior.

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 the baseline is 3. The description's transformation examples overlap with what the schema already provides and it adds no detail beyond the structured 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 (Build) and resource (CDN delivery URLs for assets) and notes the optional transformation with concrete examples. This clearly distinguishes it from mutating siblings like cloudinary_upload, cloudinary_delete, or cloudinary_move.

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 nature of the tool (you build URLs to deliver existing assets), but there is no explicit when-to-use/when-not guidance or reference to alternatives. An agent must infer that this is the delivery-URL entry point versus the management siblings.

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 safety behavior is covered structurally. The description adds only a field inventory, which is already conveyed by the output schema, and says nothing about auth, rate limits, or missing-asset behavior.

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; every item listed contributes to the scope. It is a sentence fragment lacking a verb, but the size is appropriate for the tool's simplicity.

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 a full output schema, complete parameter documentation, and annotations covering safety, the description only needs to establish purpose and scope, which it does. Minor missing context: no mention of error behavior when the public_id is not found.

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 both enum parameters are documented in the schema, so the schema does the heavy lifting. The description mentions no parameter formats (e.g. asset_id:<id> prefix) beyond what the schema already states, matching the baseline of 3.

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 the specific verb+resource implied by 'Full metadata of one asset' and enumerates the fields returned, which separates it from list/tree/url siblings. It stops short of explicitly naming a competing sibling, so it is not 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?

'one asset' implicitly signals single-asset lookup versus cloudinary_list or cloudinary_tree, but there is no explicit guidance on when to prefer this tool over cloudinary_get_url or cloudinary_download. Usage must be inferred from the noun phrase.

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, and destructiveHint=false, so the safety profile is covered. The description usefully adds that paths are root-relative and that wildcards return matches, but it says nothing about behavior on a missing folder, result ordering, or how max interacts with subfolder listings.

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 the primary scope first and the wildcard variant second; there is no filler and nothing redundant.

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 an output schema, full parameter documentation, and rich annotations, the description only needs to convey scope and the two calling modes, which it does. The remaining gap is minor: no note on cross-referencing with cloudinary_tree or on wildcard/folder error behavior.

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 every parameter, including defaults and enum values, is documented inline, so the baseline is 3. The description only restates the path/wildcard semantics the schema already explains, adding no new meaning for max, delivery_type, or resource_type.

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 and resource ("List the subfolders and assets of a Cloudinary folder") and adds the dual mode of wildcard matching, so the agent knows exactly what comes back. It does not explicitly contrast itself with the sibling cloudinary_tree, so differentiation is only implicit.

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 examples ("portfolio", "cld*", "shots/*.png") imply the shape of a call, but nothing tells the agent when to pick this over cloudinary_tree or how a folder path differs from a recursive listing. Usage is inferable rather than 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 already declare readOnlyHint=false, idempotentHint=true, openWorldHint=true, and destructiveHint=false. The description adds one genuinely useful side-effect disclosure: missing parent folders are created too. Beyond that it adds no context on permissions, path format, or behavior when the folder already exists (partly covered by idempotentHint).

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 the key side-effect front-loaded in parentheses. No wasted text.

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 an output schema present and rich annotations, the description supplies what's needed to call the tool correctly. Minor gaps remain (path format, no-op behavior on existing folders) but they are largely covered by the idempotentHint 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?

Schema coverage is 100% and the single parameter documents itself as "Folder path from the root". The description adds no syntax, separator, 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 ("Create a folder") with a parenthetical scope note. It is unambiguous and no sibling (rename_folder, move, upload) overlaps with plain folder creation, though it doesn't explicitly say so.

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 verb – you call it when you need a folder to exist – but there is no explicit when/when-not guidance or reference to alternatives such as cloudinary_rename_folder or cloudinary_move for other path operations.

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, idempotentHint=false, and openWorldHint, so the safety profile is covered. The description still adds real value by disclosing the confirmation gate and the URL-preservation behavior in dynamic folder mode, which are not visible from 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?

Three sentences, no waste, with the core action front-loaded and the confirmation mechanics trailing as supporting detail. Every sentence carries distinct information.

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 an output schema present, rich annotations, and full schema coverage, the description only needs to explain the non-obvious workflow, which it does via the confirm_token gate. Minor gap: it does not mention permission requirements or what happens to the existing asset when overwrite replaces 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?

Schema description coverage is 100%, so every parameter including dry_run, overwrite, and confirm_token is already documented in the schema. The description's 'keeps its URL unless change_public_id is set' largely restates the schema's own wording, so it adds little beyond the structured fields – baseline 3.

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 ('Move or rename an asset, or move a folder') and names the dynamic folder mode qualifier. It does not explicitly distinguish itself from the sibling cloudinary_rename_folder, which overlaps on folder operations, so an agent may still hesitate between the two.

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 useful context about the two-step confirmation flow ('first call returns needs_confirmation... confirm_token'), which tells the agent how the call sequence works. However, it never says when to prefer this over cloudinary_rename_folder or when to use dry_run instead of a real call, and offers 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?

Annotations already declare destructiveHint, idempotentHint=false and openWorldHint, so the safety profile is covered. The description adds genuinely new behavior: asset URLs are unchanged after the rename, and the first call is non-executing and returns needs_confirmation plus a confirm_token. It doesn't describe failure modes outside dynamic folder mode.

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 sentences, no filler, and the most decision-relevant facts (recursive scope, mode restriction, confirmation requirement) are all front-loaded and unmissable.

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 rich annotations and an output schema present, the description doesn't need to explain return values. It covers scope, mode restriction, and the confirmation handshake, leaving only error behavior (e.g., what happens outside dynamic folder mode) 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 100%, so all four parameters are already documented, giving a baseline of 3. The description adds the relationship between calls that explains why confirm_token exists ('runs exactly that plan') and how dry_run interacts with it, which is meaning beyond the field-level text.

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 specific verbs ('rename or move'), the exact resource ('a folder'), and the recursive scope ('with everything in it'), which is what separates it from cloudinary_move (asset-level moves). An agent can pick between them without reading 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 a hard precondition ('dynamic folder mode only') and explains the two-step confirmation flow that governs real usage. It does not explicitly exclude or route to cloudinary_move/cloudinary_mkdir, so it stops short of full when-not 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 establish readOnly, idempotent, non-destructive, open-world behavior, so the description does not need to restate safety. It usefully discloses that the result is a hierarchy including each folder's assets and that the traversal starts at root by default, but says nothing about the cost of deep/unbounded recursion or how large results are returned.

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 starting-point default is placed where the reader needs it. Nothing could be removed without losing information.

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?

An output schema exists and all four parameters are fully described, so the mechanical contract is covered. However, for a recursive tree operation the description omits practical context an agent would want — result size, traversal depth risk, and how it differs from cloudinary_list.

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 (depth, folder, max_files, resource_type) is already documented in the schema. The description only echoes the folder default, adding no extra semantics such as how depth interacts with max_files, 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?

Names a specific resource (folder hierarchy with the assets of each folder) and the starting point default, which is more than a tautology. It does not distinguish itself from the closely related sibling cloudinary_list, so an agent must infer the difference between a tree walk and a flat listing.

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 or when-not-to-use guidance is given. With siblings like cloudinary_list and cloudinary_info in the same server, the agent gets no signal for choosing between a recursive tree and a flat listing.

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 destructiveHint annotation is refined meaningfully: the description discloses that replacements are intercepted by a needs_confirmation/confirm_token handshake rather than silently overwriting, and that the resource type is inferred from the extension unless overridden. It stops short of covering directory-upload edge cases, auth requirements, or batch failure semantics.

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, front-loaded with the core action and scope, followed by the two non-obvious rules (type inference, confirmation gate). Every sentence carries distinct information with no restatement of 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?

With annotations and an output schema already present, the description supplies the one behavioural detail an agent could not infer: the confirmation-token workflow. Remaining gaps are minor — the set of 'allowed local directories' is referenced but not characterised, and multi-asset directory upload outcomes are left to the schema.

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 all eight parameters (including confirm_token, dry_run, overwrite, and the resource_type enum) are already documented in the schema. The description restates the resource_type default behaviour but adds little interpretation beyond what the schema provides, so the baseline of 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?

The description uses a specific verb ('Upload') with a precise resource boundary: a local file or directory inside allowed directories into a Cloudinary folder. It also states the resource-type inference rule, which distinguishes it from siblings like cloudinary_move, cloudinary_download, and cloudinary_delete that operate on already-stored assets.

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?

Clear context for when this tool applies (pushing local content into Cloudinary) and it explains the two-phase confirmation flow that gates destructive replacement. It does not explicitly compare against siblings such as cloudinary_move (server-side reorganisation) or cloudinary_download, so an agent must infer the boundary itself.

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