Conciseness

Dense, procedural text with no explanations of concepts Claude already knows. Minor redundancy could be trimmed: skip/`leave-onboarding` handling is spelled out in step 3, step 4, and the Rules section, and `Show me other ideas` behavior is explained in both step 2 and step 3. Not 5 because of that repetition; not 3 because padding is incidental rather than a broad over-explanation problem.

Actionability

Fully executable guidance: exact question texts ("Which one feels most useful?", "Nice! Should I build this now?"), exact option labels (`Yes, build it`, `Not now`, `Show me other ideas`), hard constraints (titles "of at most 6 words", descriptions "of at most 18 words"), and a copy-paste example option string. Specific examples cover the common conversational cases, matching the anchor 5 standard for an instruction-only skill.

Workflow Clarity

A five-step Flow with every answer branch enumerated (suggestion, `Show me other ideas`, free-text task, free-text change, skip, `answered: false`), an explicit build gate ("A build needs `Yes, build it` from the user"), and feedback loops ("Not now" returns to step 2; every card answer is read and routed). The operations are conversational and user-confirmed, so the destructive/batch validation cap does not apply.

Progressive Disclosure

Well-organized single-file structure (Input / Flow / Rules) with no dead-end or nested references, and nearly all content is needed on every invocation, so inline placement is defensible. Not 5: at ~127 lines it exceeds the simple-skill threshold, and the onboarding-specific input handling and the suggestion-format spec (title/description/apps rules) are plausible candidates for a one-level-deep reference file. Not 3: structure is clear and content is appropriately placed overall.

Specificity

Base 4: it lists several concrete actions — "Offer the three most common automations for a user as a single-choice card", "build the one they choose once they confirm, or offer more" — with only minor coverage gaps (the n8n domain is never named). The second-person phrase "from your own knowledge of their team" triggers the rubric's one-point voice penalty, reducing it to 3. Not 2: far more than minimal or generic actions are named.

Completeness

Explicitly answers both: what — "Offer the three most common automations … then build the one they choose once they confirm, or offer more" — and when — "Use it on other threads when the user asks what they could automate or wants ideas for a first workflow", plus the onboarding preload context. Matches the anchor 5 example structure with concrete trigger phrases.

Trigger Term Quality

Includes natural phrases a user would actually say: "asks what they could automate" and "wants ideas for a first workflow". Not 5: common variations such as "automation ideas", "automate tasks", or platform-specific terms are missing. Not 3: the key natural trigger phrases are present and well-matched.

Distinctiveness Conflict Risk

Clear niche (suggest-and-confirm automations, preloaded on onboarding threads) with distinct triggers. Not 5: "build the one they choose" edges into workflow-builder territory, creating minor overlap risk with a closely related build skill. Not 3: the suggestion/ideation trigger set is specific and unlikely to fire for unrelated skills.