Content Ops

Content Ops

Generates consistent content across all formats by anchoring every generation to a single source of truth: BRAND.md.

The uniformity guarantee

Content drifts when each generation re-interprets voice, terminology, and messaging from scratch. This skill prevents drift by:

1. Always reading BRAND.md first — before writing a single word
2. Routing to a sub-skill — each content type has explicit structure rules
3. Enforcing quality checks — every sub-skill has a pre-output checklist

Update BRAND.md once → every future generation inherits the change.

When to use

Use this skill whenever you are generating:

For full ad campaign strategy (audience, funnel, budget), use the ads skill instead — it has platform-specific campaign architecture. Use this skill for the copy execution within that strategy.

Required setup

Before generating any content for the first time:

1. Open BRAND.md in this directory
2. Fill out every section — especially:
   • Voice rules (always/never)
   • Terminology table (prevents off-brand word choices)
   • Primary Message and Supporting Messages (anchors every headline)
   • Proof points (prevents invented stats)
   • Compliance section (legal guardrails)
3. Save and commit BRAND.md — treat it as a controlled document

Incomplete brand manifests produce incomplete content uniformity.

Execution protocol

When asked to generate content using this skill:

Step 1 — Classify the request

Identify which content type is being requested. If unclear, ask:

• "Is this for support/help center, marketing/promotional, product page, email, or social?"
• "Which channel or platform?"

Step 2 — Read BRAND.md

Read the full BRAND.md before writing. Extract:

• Relevant voice rules for this content type
• Applicable terminology
• Proof points to draw from
• Channel-specific constraints from the Channels table
• Any compliance requirements

Step 3 — Load the sub-skill

Read the appropriate sub-skill file and apply its template and quality checklist.

Step 4 — Generate content

Produce the content using the sub-skill template structure. Flag any gaps with [NEEDS INPUT: reason] rather than inventing details.

Step 5 — Run quality checks

Before outputting, run through the sub-skill's quality checklist. Do not skip checks. Do not output content that fails a check without flagging it.

Step 6 — Output in standard format

Always structure the final response with these sections:

1. Summary — content type, channel, goal in 1-2 sentences
2. Brand anchors used — which Primary Message, proof points, and voice rules were applied
3. Content — the deliverable, in fenced blocks with content type labeled
4. Gaps / flags — any [NEEDS INPUT] items, missing proof, or compliance items to resolve
5. Variants — if multiple versions are useful (e.g., A/B subject lines), list them here

Maintaining uniformity over time

When a brand decision changes

1. Update BRAND.md — the one file
2. Do NOT patch individual content pieces — regenerate from the manifest

When content conflicts arise

If generated content feels "off brand," the fix is almost always in BRAND.md:

• Add a new "never do" voice rule
• Add the correct term to the Terminology table
• Sharpen the Primary Message

When generating a content series

State the series goal and all pieces upfront so they are generated with shared context — not one-by-one with fresh context each time. This prevents the most common source of drift: each generation starting cold.

Related skills

• ads — full campaign strategy (audience, funnel, budget, creative briefs)
• ads-copy — ad copy variations with A/B frameworks
• ads-landing — landing page audit and optimization
• annuity-life-insurance-sales — insurance-specific sales language and workflow