brand-voice-writer

name	brand-voice-writer
platforms	["cowork","claude-code"]
description	Write marketing content in TigerData's brand voice with full context on our audience, positioning, and terminology. Use this skill whenever the user asks to write blog posts, articles, landing page copy, email campaigns, social posts, one-pagers, case studies, or any marketing content. Also trigger when the user mentions brand voice, tone, writing style, ICP, target audience, or asks for on-brand writing. Even if the user just says 'write something about [topic]' without specifying brand voice, use this skill — all marketing content should be on-brand.
references	["product-marketing-context","brand-voice-guide"]

Brand Voice Writer

This skill gives you deep context on TigerData's brand, audience, positioning, and terminology so you can write marketing content that sounds like us — not generic AI output.

When to use this skill

• Any time you're writing marketing content (blog posts, articles, emails, social, landing pages, case studies, one-pagers, ad copy)
• When the user asks for "on-brand" content or mentions brand voice / tone
• When writing about TigerData products or the Timescale/PostgreSQL ecosystem
• Basically: if it's going to be published or shared externally, use this skill

SEO ambiguity check

Before starting the writing workflow, check whether the request has SEO signals that suggest it should use seo-article-writer instead. Look for:

• Keyword targeting mentioned in the request or brief ("target keyword," "rank for," "search volume")
• SERP-oriented structure (FAQ sections, comparison tables, "what is X" definitions)
• Brief mentions ranking goals or search intent
• "Alternatives" or "vs" framing
• Roundup or listicle structure comparing multiple tools/products
• The request includes an SEO brief or content brief with keyword data

If any of these signals are present but the user did NOT explicitly say "SEO article," "GEO article," or "AEO article," ask:

"This looks like it could be either a general blog post or an SEO-targeted article. Should I treat this as an SEO article (keyword-optimized, competitive verification, structured for ranking) or a standard blog post?"

If the user says SEO, direct them to invoke seo-article-writer (in Claude Code: /tigerdata-marketing-skills:seo-article-writer).

If the user says standard blog post, proceed with this skill's normal workflow.

If the request has no SEO signals, proceed normally — do not ask.

Step 0: Pre-flight check

Invoke the marketing-preflight skill via the Skill tool, passing this skill's name: field value (brand-voice-writer) as the args. Do not proceed until it completes successfully. If it stops with an error, follow its instructions and stop.

If marketing-preflight returned a stale-connector footer note, append it verbatim at the end of your final response to the user.

Also fetch the No Fly List before doing any work:

get_marketing_reference(slug: "no-fly-list")

This is a list of customers who cannot be publicly referenced. Load these names as a hard constraint: never include any No Fly List customer in any output — not as named examples, proof points, customer quotes, case study references, or any other mention. If the user requests content featuring or referencing a No Fly List customer, stop and inform them that this customer cannot be publicly referenced. If a No Fly List name appears in source material you are working from, omit it from all outputs.

Instructions

1. Load the right context

Before writing anything, fetch the reference docs declared in this skill's frontmatter.

Which docs to fetch:

• Always fetch: product-marketing-context — the what: audience, positioning, terminology, competitive landscape, proof points, and marketing strategy. This is the single source of truth for who we are, who we talk to, and what we say.
• Always fetch: brand-voice-guide — the how: detailed writing instructions per content type (web pages, blog posts, emails, slides, data sheets, social/ads), the WABL principle, absolute rules, voice principles, intro/outro guidance, and the final filter. This tells you how to actually write each format.

Fetch both docs before writing. product-marketing-context gives you the strategic context; brand-voice-guide gives you the structural playbook.

How to fetch: Use Tiger Den to fetch both docs in one call:

get_marketing_context(slugs: ["product-marketing-context", "brand-voice-guide"])

If Tiger Den is not connected, do not proceed — tell the user to run /setup. See REFERENCES.md in the plugin root for details and error handling.

2. Apply brand voice consistently

After reading the reference docs, internalize the voice and tone principles. The brand voice guide is the primary authority on how to write; the product marketing context doc is the primary authority on what to say.

Key things to watch for:

• Use approved terminology (product names, technical terms, capitalization) from the product marketing context glossary
• Match the tone and structural template for the content type from the brand voice guide (it has specific guidance for web pages, blog posts, emails, slides, data sheets, and social/ads)
• Apply the WABL principle throughout: "Would Anything [of value] Be Lost?" — if removing a sentence loses nothing, remove it
• Follow the absolute rules (no em dashes, one message per piece, lead with the problem, active voice, short paragraphs)

3. Tailor to the audience

If you know who the content is for (e.g., a specific persona from the product marketing context doc), adjust your approach:

• Lead with the pain points and motivations that matter to that persona
• Use the language and framing that resonates with their role
• Reference the use cases and scenarios relevant to their context

4. Write the content

Produce the content with the full context you've loaded. If the user hasn't specified a format, suggest one based on the topic and audience.

After writing, do a self-check against common AI writing patterns. These are the tells that most often survive a first draft, even with brand context loaded:

• Negative seesaws: "It's not X. It's Y." / "Not just X, but Y." This construction is the single most common structural AI tell. State the point directly instead of setting up a contrast.
• Forced triples: Grouping things in threes compulsively ("Innovation, collaboration, and excellence"). If there are two things, say two.
• Copula dodging: Using "serves as," "stands as," "functions as" when the sentence just means "is."
• Vocabulary cluster: Words like "delve," "landscape," "crucial," "showcase," "underscore," "tapestry," "foster," "bolster" appear far more often in AI text than human writing. Use the simpler word you'd say out loud.
• Em dashes: Our brand voice bans them entirely. Replace with commas, periods, or parentheses.
• Hallmark-card endings: "The future looks bright." "This is just the beginning." End the piece when it's done — the last paragraph should contain information, not warm feelings.

Also verify:

• Are product names and technical terms correct per the approved terminology?
• Would this resonate with the target audience?
• Is the positioning consistent with our product marketing context?

If AI patterns are still present after your self-check, or if the user wants an extra polish pass, suggest running /de-slop on the output. De-slop has a comprehensive pattern catalog and rewriting workflow purpose-built for stripping AI tells while preserving voice.

Tiger Den content enrichment

These features use Tiger Den tools that are already connected (since you fetched reference docs in Step 1).

Content search

Use search_content to find relevant existing content — published articles, case studies, and data points that can strengthen the piece. Search for topics related to what you're writing about to find supporting material.

Voice profiles

If the user asks to write in a specific person's voice (e.g., "write this like Matty would," "in Mike's style," "draft this as Jacky"):

1. Call list_voice_profiles to see who's available
2. Call get_voice_profile with the matching name to load their writing samples and voice notes
3. Apply that person's voice on top of the brand voice guide:
   • The voice profile takes precedence for sentence rhythm, tone, humor, paragraph style, and verbal tics
   • The brand voice guide and product marketing context still govern terminology, absolute rules (no em dashes), and positioning
   • Think of it as: the brand voice guide sets the floor and guardrails, the voice profile sets the personality

If the user doesn't mention a specific person, don't load a voice profile — just use the brand voice guide as the default voice.