Category Narrative Map
Reveal the category stories, competitor messaging, and shifts shaping your position.
The skill explicitly treats fetched content as untrusted data, requires robots.txt preflight, labels proxy signals, prevents competitor assertions from becoming facts, and asks for confirmation before memory writes or claim promotion; however, Firecrawl/Tavily transmit target URLs and queries to third parties, while the actual registry authority boundary is delegated to external protocols. The skill itself lacks complete data-flow, rollback, and sensitive-data handling details, so points are deducted.
The contract, numbered procedure, done criteria, and termination rules are broadly consistent, with clear Measured/User-provided/Estimated/proxy distinctions; however, the key path depends on several connectors, archive services, and shared protocols without skill-specific tests, abnormal-input behavior, or diagnosable failure messages. The static ceiling is applied.
Triggers, inputs, output components, competitor-set handling, adjacent-category coverage, and non-fit boundaries are clearly stated, with Chinese labeling and descriptions; however, core drift analysis depends on reachable scraping and Wayback services. Mainland-China availability is not established, and proxy or pasted-data fallbacks reduce coverage, so points are deducted.
The file provides frontmatter, version, license, host compatibility, Quick Starts, an input/output contract, sources, procedure, save confirmation, handoff, and next-skill guidance; however, maintenance ownership, update path, changelog, dependency versions, and troubleshooting are not clear in the skill itself, and many key controls are externalized to referenced files not supplied here.
The intended output covers category stories, language conventions, per-competitor teardown, messaging drift, and handoff, which could directly support positioning work; however, static evidence cannot establish that scraping, archive comparison, or document production succeeds on real inputs. Results depend heavily on competitor URLs, archive availability, and user-supplied context, so the static ceiling applies and points are deducted.
The skill requires source-type labels, URLs for Measured rows, and explicit separation of proxy signals from facts, providing some auditability; however, there are no skill-specific tests, third-party execution records, or documented multi-source reproductions. Repository tests mainly cover shared architecture and audit validation, not this skill's key paths, so points are deducted.
- Scraping and search may send target URLs and queries to third parties such as Firecrawl/Tavily; do not provide confidential competitive leads or unpublished material.
- If Wayback or scraping services are unavailable, drift analysis cannot meet the done criteria; mark the evidence missing instead of inferring historical changes.
- The skill asks for confirmation before memory persistence while its done condition refers to saving; treat explicit user confirmation as mandatory before any durable write.
- No skill-specific tests or execution records are provided, so the score does not establish runtime reliability or accuracy of marketing conclusions.
What does this skill do, and when should you use it?
Category Narrative Mapper researches the narrative terrain of a market as the second move in the TALE Trace phase. It identifies dominant category stories, points of view, recurring language, framing clichés, and saturated terms, then breaks down each named competitor's arc, claimed onlyness, proof pattern, and primary frame. It also compares current competitor copy with dated archive snapshots to identify messaging drift. The skill produces a category narrative map; it does not build the positioning canvas, capture audience beliefs, adjudicate claims, or calculate a TALE result.
Confirms the category and competitor set, reuses prior competitor-analysis findings when available, and gathers public messaging through the repository's Firecrawl connector, Tavily search signals labeled proxy, and Wayback archive connector. It names two to four dominant category stories and catalogs recurring vocabulary, framing clichés, table-stakes language, and saturated language. For every named competitor, it records the narrative arc, claimed onlyness sentence, proof pattern, and primary framing, then adds a current-versus-archived messaging-drift note with as-of dates. Outputs are labeled Measured, User-provided, or Estimated and are written to memory/narrative/category-narrative-mapper/ only after the user confirms saving.
- A brand team is developing a new narrative and needs to understand the frames already used by the category.
- A market researcher needs a structured teardown of how several competitors tell their stories and prove them.
- A positioning lead wants to compare a competitor's current homepage messaging with archived versions.
- A content strategist needs to distinguish must-speak category language from overused framing clichés.
- A product marketing team needs category-frame candidates and named alternatives for downstream positioning work.
What are this skill's strengths and limitations?
- Produces a focused map covering category stories, language conventions, competitor narratives, and messaging drift.
- Separates Measured, User-provided, and Estimated evidence and labels search signals as proxy.
- Supports current-copy versus Wayback archive comparisons.
- Defines clear scope boundaries and hands off to downstream positioning work.
- The repository is Apache-2.0 licensed and supports Claude Code plus compatible Agent Skills hosts.
- Requires public competitor pages or user-provided material; closed social platforms are limited to pasted excerpts or proxy reads.
- Does not build the positioning canvas, adjudicate claims, or compute the TALE profile result.
- Scraping and archive comparisons depend on the Firecrawl, Tavily, and Wayback connectors; the supplied material does not provide per-connector test results.
- Persistent memory writes require user confirmation.
- No standalone test suite for this individual skill is shown in the source material.
How do you install this skill?
Clone the repository with git clone https://github.com/aaron-he-zhu/aaron-marketing-skills. In Claude Code, run /plugin marketplace add aaron-he-zhu/aaron-marketing-skills and then /plugin install aaron-marketing@aaron. The README also supports npx skills add aaron-he-zhu/aaron-marketing-skills for Agent Skills-compatible hosts. The supplied material gives a single-skill installation example but does not show this skill's exact standalone command.
How do you use this skill?
Use a prompt such as “Map the category narrative for [product/category]. Competitors: [names]” or “Tear down how [Competitor A], [Competitor B] tell their story — arc, claimed onlyness, proof pattern.” For historical comparison, use “Show how [competitor]'s messaging has shifted over the last 2 years — scrape their site now and compare against the archive.” Provide the category, product, and competitor names or URLs.