write-workflow-as-code

MUST use when writing or modifying Windmill Workflow-as-Code scripts using workflow, task, step, sleep, approvals, taskScript, taskFlow, task_script, or task_flow.

Quality

83%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Medium

Suggest reviewing before use

Quality

Content

77%

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

A highly actionable, well-sequenced guide with strong code examples and validation checkpoints, weakened mainly by repeated deploy cautions and a large inline API-reference block that should be factored into a separate reference file.

Suggestions

Move the TypeScript and Python API signature blocks (lines ~244–534) into a separate `references/api.md` and keep only a concise import + key-usage snippet in SKILL.md, linking out for full signatures.

State the "only deploy when explicitly asked; preview for testing" rule once in a dedicated section and remove the three other repetitions to cut redundancy.

Add a short "Quick start" section near the top that shows the minimal write → generate-metadata → preview loop before the detailed checkpoint/error-handling material.

Dimension	Reasoning	Score
Conciseness	The body is dense and largely assumes Claude's competence, but the "only deploy when explicitly asked" caution is repeated roughly four times (e.g. lines 15, 19, 27, 45) and the full TS+Python API signature blocks run ~290 lines, so it could be tightened — anchor 2 rather than the lean anchor 3.	2 / 3
Actionability	It provides executable TypeScript and Python code, exact non-aliased SDK import names, and concrete CLI commands (e.g. `wmill script preview`, `wmill generate-metadata --dry-run`), matching the copy-paste-ready anchor 3.	3 / 3
Workflow Clarity	Steps are clearly sequenced with intent-based decision rules (preview vs run vs sync push) and explicit verification checkpoints (preview-before-deploy, `--dry-run`, diff regenerated locks, pin versions), matching the anchor-3 validation-checkpoint bar.	3 / 3
Progressive Disclosure	The skill is a single ~530-line monolith with the large inline API-reference block being content that should be split into a separate reference file, and no bundle references exist — anchor 2 (structure present but content that should be separate is inline).	2 / 3
	Total	10 / 12 Passed

Description

90%

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, distinctive description with an explicit trigger clause and good natural keyword coverage; its only weakness is that the stated actions (writing/modifying) are generic rather than a comprehensive list of concrete capabilities.

Dimension	Reasoning	Score
Specificity	"writing or modifying Windmill Workflow-as-Code scripts" names the domain and two actions but is not a comprehensive list of concrete capabilities, so it sits at the anchor-2 level rather than the multi-action anchor 3.	2 / 3
Completeness	It states both what ("writing or modifying Windmill Workflow-as-Code scripts") and when ("MUST use when ... using ..."), with an explicit trigger clause, satisfying the anchor-3 both-what-and-when requirement.	3 / 3
Trigger Term Quality	It lists "workflow, task, step, sleep, approvals, taskScript, taskFlow, task_script, task_flow" — good coverage of natural terms a WAC user would actually say, matching the anchor-3 natural-keyword bar despite a few technical entries.	3 / 3
Distinctiveness Conflict Risk	"Windmill Workflow-as-Code" with its specific SDK trigger terms is a clear niche unlikely to fire for unrelated skills, matching the anchor-3 distinct-trigger bar.	3 / 3
	Total	11 / 12 Passed

Validation

93%

Warnings & errors only

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

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (535 lines); consider splitting into references/ and linking	Warning

	Total	15 / 16 Passed

Repository: windmill-labs/windmill
Path: system_prompts/auto-generated/skills/write-workflow-as-code/SKILL.md
Commit: 33d2845

Reviewed: 2 days ago

Table of Contents

Discovery Implementation Validation

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.