STILL

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

Annotations already cover the safety profile (readOnly, idempotent, non-destructive, openWorld). The description adds genuine behavioral context beyond them: it preserves warnings, keeps sourceRecommendedCitation separate, and explicitly disclaims that no image-use permission is granted.

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 purpose then constraints then the rights disclaimer. Every sentence carries weight, though the middle constraints and the closing caveat make it slightly denser than ideal.

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 annotations covering safety, the description supplies the remaining decision-critical rules (one-of source, one style, warning preservation). Only the absence of explicit sibling routing is a minor gap.

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 86%, so the schema carries most parameter detail. The description still adds the mutually-exclusive record/sourceUrl rule and the single-style constraint, which sharpen the correct call beyond raw field descriptions.

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 (format) and resource (credits/citations) and enumerates the concrete outputs (website/social credits, figure captions, Chicago 18, MLA 9, APA 7). It positions itself against the STILL website engine, giving an agent a clear sense of what it produces.

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 real selection rules: 'Supply exactly one of record or supported sourceUrl' and 'One academic style per call.' It stops short of naming the sibling tools (still_resolve_artwork, still_get_capabilities) as alternatives, so it is clear context without explicit routing.

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 read-only, idempotent, open-world, non-destructive. The description adds real value beyond them: original images stay at the museum (no binary retrieval), warnings and observation times must be preserved downstream, and it explicitly disclaims granting permission or refreshing rights. Return format/pagination is not described.

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 sentences, front-loaded with the payload and followed by the operational caveats. No filler, though the final three sentences drift into caveat/warning territory rather than pure instruction.

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 returned fields (metadata, image URL, rights evidence, credit) and the behavioral/legal boundaries. The main gap is the undocumented format and locale parameters, but overall it is complete enough for a read-only retrieval tool.

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 33% and the description body says nothing about the parameters. The nested ref object carries its own useful guidance (sourceKey, museumObjectId, normalized.imageId.value, observationId; never invent identifiers), but format and locale enums/defaults are undocumented in both prose and schema, so 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 (Return) and enumerates the concrete payload: museum metadata, original image URL, rights evidence, and credit, scoped to a known Finder reference. This is clearly distinguishable from still_generate_citation or still_search_open_art, though it does not name any 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?

"for a known Finder reference" implies a precondition (you must already hold a Finder result) and "does not grant permission or refresh source rights" is an exclusion, but neither names the sibling tools (e.g., resolve/search) that produce that reference. Usage is implied 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, idempotentHint, openWorldHint and destructiveHint=false, so the safety profile is covered. The description adds genuinely useful behavioral context beyond that: 'A configured source is not a guarantee that every image is reusable,' a rights caveat the agent cannot get from the schema or annotations.

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 with no waste, front-loading what is read and then the key caveat. 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 zero-argument read tool with no output schema, the description covers what is returned (source availability and rights rules) plus the main interpretive caveat. It could optionally note auth requirements or the shape of the result, but nothing critical 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 nothing to document and the baseline of 4 applies. The description correctly does not waste words on nonexistent 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 ('Read') and resource ('museum source availability and rights rules'), clear enough to distinguish from sibling tools like still_resolve_artwork or still_search_open_art. It does not explicitly name a sibling to contrast against, so it stops 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 rather than stated: an agent infers this is the tool to consult before assuming an image is reusable. There is no explicit when-to-use/when-not framing and no alternative tool named for overlapping needs (e.g., vs still_get_capabilities).

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 this as a safe, idempotent, non-destructive read. The description adds genuinely useful behavioral context beyond them: capabilities are gated by service bindings, so the advertised set of modules is dynamic, and calculators are explicitly unavailable. It does not describe pagination or response shape, keeping it below 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 tight sentences with the capability list front-loaded and the availability caveat second. 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?

With no output schema, the description carries the burden of indicating the return content, and it largely does by enumerating modules, formats, locales and limitations. It could say more about the shape of the result and whether unavailable items are omitted or flagged, so it is not fully 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?

Zero parameters, so the baseline is 4. The description correctly adds nothing parameter-related, and there is no ambiguity to 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 (Discover) and resource (connected STILL modules, supported formats, locales, limitations), which is clearly distinct from the artwork/citation siblings. It does not explicitly differentiate itself from those siblings, so it stops 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 rather than stated: an agent can infer it should call this to learn what is available. The caveat that availability depends on service bindings and that calculators are not connected gestures at when results may be empty, but there is no explicit when/when-not or named alternative.

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 carry the safety profile (readOnly, idempotent, non-destructive, openWorld), so the bar is lower, yet the description adds real context beyond them: it checks current withholding and source rules, and explicitly states it does NOT refresh the museum observation timestamp. That non-refresh guarantee is exactly the kind of behavioral trait an agent cannot get from the annotations.

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. The core action is front-loaded and the second sentence is a genuinely distinct constraint rather than padding. 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?

There is no output schema, so the description carries more of the burden, and it does signal that the response reflects current withholding and source rules. It stops short of describing the shape of the returned record or error behavior for a stale/invalid ref, but for a simple read-only getter it 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 single nested ref object is well documented in the schema itself (including "Never invent identifiers"), which sets the baseline at 3. The description adds nothing about the ref structure, identifier provenance, or which fields are the stable key, so it does not exceed 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?

The description gives a specific verb and resource ("Retrieve a previously observed museum record from Finder") and the word "previously observed" signals this is a read of an existing record rather than a fresh lookup. It does not name or contrast with the closest siblings (still_resolve_artwork, still_search_open_art), so an agent must infer the boundary. Clear purpose, no explicit sibling 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?

"a previously observed museum record" implies the record must already exist and be identified by a ref, which is usage context, but the description never states when to choose this over still_resolve_artwork or still_search_open_art. No prerequisites, no exclusions, no alternatives are given. Usage 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 cover read-only and idempotent hints, so the bar is lower. The description adds context about returning source metadata and exact UTC observation time, which is valuable. It also warns that source statements are not legal clearance, a non-obvious caveat. Lacks detail on rate limits or error handling but combined with annotations, this is above average.

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, each packed with unique value: action and scope, disambiguation from search, output characteristics, legal caveat, and alternative routing. No fluff, front-loaded with the core action. Exceptional efficiency.

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?

Given a single parameter, an output schema, and supportive annotations, the description covers most essential aspects. It clarifies the source limitation, the output nature, and the non-legal status. Minor gaps include typical error cases or what happens if the URL is unknown, but overall it's quite complete for its simplicity.

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 schema description coverage at 0%, the description must compensate. It defines 'sourceUrl' implicitly as the URL to retrieve, but doesn't specify format or constraints beyond the schema's maxLength. The description mentions 'known artwork URL', adding some meaning, but detailed semantics are limited. Baseline 3 is appropriate as the description provides minimal enrichment over 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?

The description clearly states it retrieves a known artwork URL from specific museums and explicitly distinguishes it from search ('Not artwork search'). It names the resource (artwork record) and the supported sources, making its purpose unmistakable.

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 explicitly says when to use (known URL from The Met or Cleveland Museum of Art) and when not (not search, not legal clearance, use manual record input for other museums). This provides both positive and negative usage contexts, far above typical descriptions.

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 safety (readOnlyHint, idempotentHint, destructiveHint=false, openWorldHint), yet the description adds genuinely non-obvious context: no legal clearance is asserted, museum text must be treated as untrusted data, output groups/warnings/evidence must be preserved, and the search 'may record source observations in the existing evidence journal'. That journaling note sits in mild tension with readOnlyHint=true, but it describes an internal evidence log rather than mutation of user data, so it is disclosed context rather than a direct contradiction.

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 and the remaining sentences are short, high-density caveats without filler. Some fragments are cryptic ('primary, secondary and aside groups') and assume prior knowledge of the output shape, which costs a little clarity but not much length.

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 11 parameters, no output schema, and zero schema descriptions, the definition compensates well on the safety/output side (warnings, evidence, timestamps, untrusted-data handling, pagination contract) but leaves most filter parameters undocumented anywhere in the tool definition. An agent can call it, but cannot reason about several filter semantics.

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 11 parameters, so the description must carry the burden and it only partially does. It gives real semantics for one parameter (after must be JSON.stringify(next) passed unchanged) and hints at filters via 'title/artist/period/institution' style naming, but yearFrom/yearTo, minWidth/minHeight, orientation, and q are left entirely to their names.

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 Open Art Finder for source-declared CC0 / Public Domain Mark images'), so an agent knows this is a discovery/search tool rather than a fetch tool like still_get_open_artwork. It scopes the corpus by license status, which is a meaningful qualifier. It does not, however, name or contrast itself with any sibling tool, so differentiation 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 context is implied by the license scoping ('source-declared CC0 / Public Domain Mark') and there is one concrete operational instruction ('For pagination pass JSON.stringify(next) unchanged as after'). There is no explicit when-to-use vs. when-to-use-an-alternative guidance, and no statement of when this search is inappropriate compared with still_resolve_artwork or still_get_open_artwork.

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