CtrlK
BlogDocsLog inGet started
Tessl Logo

camunda-api-zod-schemas

Use when you add, change, or remove schemas, fields, enums, filters, sort fields, or endpoints in @camunda/camunda-api-zod-schemas (webapp/client/packages/camunda-api-zod-schemas/); when you align the package with the OpenAPI spec in zeebe/gateway-protocol/src/main/proto/v2/; when you add a module or a release-line version tree; or when you increase the package version and prepare a release.

72

Quality

89%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

92%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

An exemplary operational skill: lean, fully executable, and clearly sequenced with real validation and error-recovery loops around an irreversible publish step. Its single real defect is that the four references to reference.md point at a file missing from the bundle, leaving the mapping tables and changelog template unavailable.

Suggestions

Ship reference.md in the skill bundle (e.g. references/reference.md) containing the "Spec to Zod" mapping table, the "Names" table, and the changelog template that Procedures 3, 4, and 7 depend on, and update the four links to its actual path.

Alternatively, if the tables are short, inline the essential mappings (e.g. the handful of most common spec-to-Zod conversions) directly in the relevant procedures rather than pointing to an absent file.

DimensionReasoningScore

Conciseness

Lean imperative writing throughout ("The spec is the reference. Do not add a field, enum value, or endpoint that is not in the spec.") with no explanation of concepts Claude already knows (Zod, REST, npm, builds are never introduced). Version numbers (8.10/8.11, 0.0.93→0.0.94) are operational facts of this version-tree workflow, and the issue-#39752 note is a necessary currency guard, so no padding penalty applies.

5 / 5

Actionability

Fully executable guidance throughout: exact commands ("grep -n \"agent-instances\" zeebe/gateway-protocol/src/main/proto/v2/rest-api.yaml", "npm run build -w @camunda/camunda-api-zod-schemas", "gh workflow run publish-zod-schemas.yml --repo camunda/camunda --ref main -f dry_run=false"), exact file paths, a real PR diff, and copy-paste TypeScript/JSON snippets for endpoints, vite entries, and package exports.

5 / 5

Workflow Clarity

Nine explicitly sequenced procedures with numbered cross-references ("Do Procedure 7 and Procedure 8"). Procedure 8 is a dedicated validation pass (build, package typecheck, workspace typecheck, format, lint) with error-recovery feedback ("If a command shows errors, find the cause and remove it"; "If a consumer type error occurs, change the consumer code or the MSW mocks"). The irreversible publish is gated by an explicit safety checkpoint ("Do not start the workflow yourself") plus post-publish verification via "npm view @camunda/camunda-api-zod-schemas version" — so the destructive-operation cap does not apply.

5 / 5

Progressive Disclosure

The body is well organized (terms table, rules, numbered procedures, examples table) and its references to [reference.md](reference.md) are clearly signaled and one level deep. However, the actual bundle contains no reference.md (no references/ directory at all), so the offloaded "Spec to Zod" mapping table, "Names" table, and changelog template are unreachable — breaking navigation to content the procedures depend on. This is better than anchor 2 (nothing is inlined or buried) but the dangling reference caps it below anchor 4's "references mostly clear" placement.

3 / 5

Total

18

/

20

Passed

Description

86%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, trigger-rich description that pins a named package, its location, and a comprehensive set of concrete change operations. Its main gaps are the implicit "what" (capabilities only appear inside "Use when..." clauses) and the second-person phrasing the rubric penalizes.

Suggestions

Open with a third-person capability statement before the triggers, e.g. "Guides changes to the Zod schemas in @camunda/camunda-api-zod-schemas... Use when...", to make the "what" explicit and avoid the second-person penalty.

Drop "you" from the trigger clauses: "Use when adding, changing, or removing schemas, fields, ..." preserves all trigger terms in third-person voice.

DimensionReasoningScore

Specificity

Enumerates comprehensive concrete operations ("add, change, or remove schemas, fields, enums, filters, sort fields, or endpoints", "align the package with the OpenAPI spec", "add a module or a release-line version tree", "increase the package version and prepare a release"), matching the anchor-5 example; however, the rubric penalizes second-person voice ("Use when you add...") by one point on this dimension, lowering it to 4.

4 / 5

Completeness

The "when" is explicit and multi-condition (anchor-5 quality), but the "what" exists only implicitly inside the trigger clauses — no statement of what the skill provides (procedures/rules for changing the package). It sits between anchors 4 and 5: stronger than anchor 4 (whose weakness is an under-explicit "when") but not clearly answering "what" the way anchor 5 requires.

4 / 5

Trigger Term Quality

Comprehensive natural trigger terms for the niche: the package name "@camunda/camunda-api-zod-schemas", "schemas", "fields", "enums", "filters", "sort fields", "endpoints", "OpenAPI spec", "module", "release-line version tree", "package version", and "release", with add/change/remove synonyms covering variations.

5 / 5

Distinctiveness Conflict Risk

Names a specific package, its monorepo path ("webapp/client/packages/camunda-api-zod-schemas/"), and the spec directory — a clear niche with distinct triggers and minimal risk of firing for an unrelated skill.

5 / 5

Total

18

/

20

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

relative_links

Relative link issues: 4 missing

Warning

Total

15

/

16

Passed

Repository
camunda/camunda
Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.