audience-generate

Generate a person audience from a plain-English brief — discover the signals behind it, score them with the math visible, and compose them into an audience whose reach is measured. The build step behind /watt:audience; routes to the leaf that fits how the audience should be built. Produces a signal stack + measured reach, or a roster — never an export, never contact data. Not a user command — /watt:audience is the front door. Use when a build-shaped ask arrives — "build me an audience of …", "I need about N people who …", "the highest-intent people in-market for X", "where do these people cluster", "a lead list of in-market companies", "match my customer list / build from my customers" — or when a /watt:explore signal pool is ready to become an audience.

Quality

86%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

73%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.

A well-structured, action-oriented routing skill with an explicit sequenced spine and strong validation discipline. Its main weakness is verbosity — long parenthetical sentences throughout the body could be tightened to improve token efficiency.

Suggestions

Tighten the dense parenthetical asides in the routing bullets (audience-generate-search / audience-generate-list) — split into shorter sentences or move detail to the referenced leaf files.

Replace the `<Bundle root>/context/*.md` placeholder paths with concrete relative paths so the one-level-deep references are verifiable and copy-paste ready.

Consider moving the language translation table and the record-file format examples into a referenced file to keep SKILL.md a lean overview.

Dimension	Reasoning	Score
Conciseness	Substantive and domain-specific rather than padded with concepts Claude already knows, but the prose is dense with very long sentences and heavy parenthetical asides (e.g., the routing bullet descriptions) that could be tightened without losing meaning.	3 / 5
Actionability	Gives concrete, executable guidance — exact `cat "<Bundle root>/context/discovery.md"` commands, a named script (`scripts/signal_profile.py`), explicit record-file formats and column lists — with only minor gaps in copy-paste readiness.	4 / 5
Workflow Clarity	The shared spine is a clearly sequenced A–E procedure with explicit validation checkpoints ("Nothing is composed unseen", "End every turn at its question", re-show + re-write on every working-set change) and feedback loops (pivot freely, failure-modes section), satisfying the anchor-5 bar.	5 / 5
Progressive Disclosure	Detailed procedures (discovery, scoring, adjacencies, record, signal-metrics) are referenced out one level deep and clearly signaled via `cat` reads, but the references use a `<Bundle root>` placeholder rather than concrete relative paths and some inlined material (language table, record examples) could live in referenced files.	4 / 5
	Total	16 / 20 Passed

Description

95%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, third-person description that crisply states capabilities, gives rich natural trigger phrases, and cleanly separates itself from sibling skills. Slightly below a perfect specificity score only because some routing phrasing dilutes the concrete action list.

Dimension	Reasoning	Score
Specificity	Lists several concrete actions — "discover the signals behind it, score them with the math visible, and compose them into an audience whose reach is measured" plus "Produces a signal stack + measured reach, or a roster" — but interleaves process phrasing ("routes to the leaf that fits") that keeps it just shy of fully comprehensive coverage.	4 / 5
Completeness	Explicitly answers both what (generate/score/compose/measure an audience, producing a signal stack + reach or roster) and when ("Use when a build-shaped ask arrives" with concrete trigger phrases), matching the anchor-5 pattern.	5 / 5
Trigger Term Quality	Embeds a comprehensive set of natural trigger phrases users would actually say — "build me an audience of …", "I need about N people who …", "the highest-intent people in-market for X", "match my customer list / build from my customers" — covering synonyms and varied phrasings.	5 / 5
Distinctiveness Conflict Risk	Carves a clear niche (person-audience build behind /watt:audience) and explicitly distinguishes itself from exports, contact data, and the analyze/activate steps ("never an export, never contact data"), minimizing wrong-skill triggering.	5 / 5
	Total	19 / 20 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
referenced_paths_exist	Referenced path issues: 2 missing	Warning

	Total	15 / 16 Passed

Repository: wattdata/plugin
Path: skills/audience-generate/SKILL.md
Commit: 880e260

Reviewed: 4 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.