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 real behavioral detail beyond that: a reason is returned per change, and a warning is emitted when a token is removed without prior deprecation — information an agent could not infer 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 sentences, zero filler, and the output contract is front-loaded before the usage hint. Every clause carries new 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, the description needn't restate return values, yet it still summarizes the bump, per-change reasons and warnings. Combined with a fully documented single parameter, an agent has everything needed to call it correctly.

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

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

Schema coverage is 100%, so the baseline is 3, but the description goes further by enumerating the five change kinds (added, removed, renamed, value-changed, deprecated) that the `changes` array accepts, giving the agent the semantic vocabulary of the required input.

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 — take token changes and return the required semver bump (major/minor/patch) with per-change reasons and warnings. That output shape makes it clearly distinct from siblings like lint_naming or audit_css, though it never names an alternative 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?

It gives a clear context for use: 'Use it to prepare a release or review a change set.' That is enough for an agent to know when it applies, but there are no exclusions or named alternative tools for adjacent tasks such as linting or proposing changes.

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 genuine behavioral context beyond that: the categories of findings it emits and the ok=false contract when at least one error is present. It does not describe pagination or result ordering, but for a non-destructive, idempotent audit that is minor.

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 filler, front-loaded with the action and the enumerated findings, then the usage cue. Every clause carries information an agent needs for selection.

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 100% covered schema and an output schema present, the description only needs to cover purpose and behavior, which it does. It could be slightly stronger by clarifying how suggestions are scoped by mode/brand, but nothing essential for correct invocation 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 description coverage is 100%, so css, mode, brand, and allow are already documented in the schema. The description adds no semantics for mode/brand scoping or the allow-list behavior beyond what the schema provides, 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 (audit) and resource (CSS snippet) and enumerates exactly what it reports: hard-coded colors/dimensions with replacement tokens, deprecated token usage, and nonexistent token references. This distinguishes it clearly from siblings such as lint_naming, search_tokens, and find_token_for_value.

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 it to review generated or legacy CSS" gives clear when-to-use context. It does not name alternatives or state when-not to use it (e.g., versus lint_naming), so it falls just short of explicit routing guidance.

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

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

Annotations already establish the read-only, idempotent, closed-world safety profile, and the description adds real behavioral detail beyond them: the result ordering (exact → semantic → primitive) and the fuzzy fallback to perceptually nearest colors or dimensions. It never states implied limits on the fallback (e.g. how distant a match can be) or whether approximate results are flagged.

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, with the matching strategy front-loaded ahead of the usage cue. The parenthetical format examples mildly duplicate the schema's value description but are defensible as quick-scannable input hints.

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 shape needn't be explained, and the description covers the input formats and result ranking that matter for correct invocation. Gaps remain around how brand/mode/property constrain matching and how the limit interacts with the fuzzy fallback tier.

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 83%, so the schema already documents mode, type, brand, property and limit. The description only reinforces the value formats, which are also examples in the schema, and its ranking language hints at property-as-intent without making that explicit. It adds little beyond the baseline for the five optional 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 pair: given a hard-coded literal, return the tokens carrying it, with an explicit match strategy (exact first, semantic before primitive, then nearest). Input formats are enumerated. It doesn't name sibling tools like search_tokens or resolve_token, so an agent must infer the boundary itself.

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 it whenever code or a design contains a literal value, to replace it with a token" gives a clear activating condition. There are no exclusions and no named alternatives, so it stops short of the when/when-not/alternative guidance that would earn 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 safety (readOnlyHint, idempotentHint, destructiveHint=false), so the bar is lower. The description adds real behavioral value beyond that: it discloses the rule set being enforced and that the result is a violation list plus a derived corrected name 'when one can be derived', which is a meaningful caveat about non-determinism of the output.

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?

Front-loaded with the tool's purpose, then packs the grammar rules densely without filler. The rule enumeration is long but each clause (kebab case, intent words, state last, no brand/mode) carries distinct enforcement information, so nothing is wasted.

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

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

An output schema exists, so return values need not be re-explained, and annotations cover the safety profile. Combined with 100% parameter schema coverage, the description supplies everything an agent needs about intent, rules, timing, and result shape.

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?

Only one parameter, and schema description coverage is 100% — the schema already documents 'names' as dot-path strings with examples. The description's grammar text implies what valid input looks like but adds no parameter-specific syntax or formatting detail. 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 verb and resource ('Check proposed token names') and enumerates the exact grammar being enforced (category.role.variant[.state], lowercase kebab, state last, no brand/mode). It does not explicitly name or contrast the other siblings such as audit_css or find_token_for_value, though 'Call it before proposing a token' does implicitly position it against propose_token.

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 clear timing guidance: 'Call it before proposing a token.' That tells the agent when in the workflow to invoke it. It omits when-not-to-use and doesn't address the other siblings (e.g., when to lint vs. audit_css), so it falls short of a full 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 mark this destructive (readOnlyHint=false, destructiveHint=true, idempotentHint=false), and the description adds the crucial guardrail context beyond them: nothing is written unless confirm is true and the server was launched with writes enabled, plus the exact validation sequence and what the preview contains. It also tells the agent the intended workflow (preview, then human review), which annotations alone cannot convey.

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 purpose, then the validation pipeline, then the write gate, then the recommended workflow. Every sentence carries information an agent needs and none repeats the schema or title verbatim.

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?

Despite having 7 parameters, an output schema and rich annotations, the description still covers the gaps that matter: alias typing rules, the confirm/writes-allowed gate, and how to treat the preview. Nothing needed to invoke it correctly is missing; return-shape detail is rightly left to the output 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 already 86%, so baseline is 3, but the description adds real meaning: value must be an alias that exists and shares the same type, and confirm is the hard gate for the write. It stays silent on brand/mode/description semantics, which the schema already covers.

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 ('Prepare the addition of a token') plus the resource and the exact scope of work: lint_naming validation, collision/scope/value checks, and a preview return of target file, JSON Patch and unified diff. This clearly distinguishes it from siblings like lint_naming (a validation used internally) and search_tokens/resolve_token (read-side 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?

Gives clear operating context ('Default to a preview and let a human review it') and explains the condition that turns it into a write (confirm true AND server started with writes allowed). It does not explicitly name a sibling alternative or spell out when-not-to-use, so it falls short of the top tier but leaves little to infer.

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 openWorldHint=false, so safety and side-effect behavior are covered. The description adds meaningful context beyond annotations by disclosing that the result includes the alias chain that produced the final value, which helps an agent understand the return shape without relying solely on the output schema.

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, front-loaded with the core action and scoped by brand and mode, with a compact inline example. 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?

Inputs are fully described in the schema, an output schema exists, and the description explains the resolution result including the alias chain. An agent has enough to call and interpret the tool correctly.

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

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

Schema description coverage is 100%, and the schema already documents accepted token reference forms (dot path, custom property, var()) and example brand/mode values. The description largely repeats what the schema states, adding only the dark-mode use case, so baseline 3 is appropriate.

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: 'Get the final value of one token for a given brand and mode, with the alias chain.' The example (color.fg.default -> color.neutral.900) makes the output concrete and clearly distinguishes it from sibling search_tokens or find_token_for_value.

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 clear contexts for use: 'to explain a value' and 'to check what a token becomes in dark mode.' It does not explicitly name alternatives or exclusions, but the stated scenarios are sufficient to guide selection among siblings for most cases.

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 genuinely new behavior: ranking rules (semantic above primitives, deprecated last) and the returned fields per brand and mode, which an agent cannot infer from the schema.

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 purpose then the usage imperative, then return/ranking behavior. No filler, though the ranking clause and return-field list are slightly dense for one sentence.

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 spelled out, yet the description adds ranking semantics that matter for interpreting results. Only the undocumented 'limit' parameter and pagination/truncation behavior are 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?

Schema coverage is 75%; query, brand and mode are documented in the schema, though 'limit' has no description at all. The description reinforces the query style and implies brand/mode shape the resolved value, but adds no format or defaulting detail 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 ('Find design tokens by intent or name') with three concrete example queries, which makes the search semantics unambiguous. It does not name a sibling to contrast with, so the boundary with find_token_for_value (reverse value lookup) or resolve_token is left 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?

'Use it before writing any color, spacing, radius or font value' gives a clear trigger condition tied to the agent's task. There are no explicit exclusions or named alternatives for cases like exact-name resolution, 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.