content-brief

name: content-brief description: Generate a platform-specific content brief for a pillar with built-in guardrails from corrections log, wins log, performance data, and Q2 strategic goals. Produces ready-to-hand-off briefs for the creative team.

Content Brief Generator

Produces a content brief for a specific pillar and platform. Pulls from performance data, client intelligence (corrections, wins, goals), and platform knowledge to generate a brief with built-in guardrails.

Trigger

User says: "content brief for [pillar]", "brief me on [pillar] for [platform]", "create a brief", "what should we post for [pillar]?", or any request for content direction.

Step 0: Load Knowledge

1. knowledge/format-playbook.md — Recommended formats per pillar, format performance data
2. knowledge/platform-behavior.md — Platform-specific best practices
3. knowledge/engagement-taxonomy.md — What engagement types to optimize for
4. knowledge/metric-glossary.md — Metric context for performance-driven recommendations
5. docs/synchrony/q2-2026-pillar-goals.md — Client strategic direction per pillar
6. docs/synchrony/signals/corrections.md — What NOT to do (language, positioning, creative guardrails)
7. docs/synchrony/signals/wins.md — What HAS worked (reinforcement signals)

Step 1: Determine Scope

1. Which pillar? Map to canonical campaign slug. If not specified, ask.
2. Which platform(s)? Determine target surface. If not specified, recommend based on format-playbook.md pillar-to-format mapping, or ask.
3. Which month? For the content calendar. Default to next upcoming month.
4. How many posts? If not specified, reference current cadence from database.

Step 2: Pull Performance Context

2a. Historical performance for this pillar

SELECT p.platform, avg(p.engagement_rate_pct) as avg_er,
       avg(p.total_engagements) as avg_eng,
       count(*) as post_count
FROM posts p
WHERE p.campaign_id = (SELECT id FROM campaigns WHERE slug = $1)
GROUP BY p.platform
ORDER BY avg_er DESC;

2b. Top performing posts for this pillar

SELECT p.*, rp.name as period_name
FROM posts p
JOIN reporting_periods rp ON p.reporting_period_id = rp.id
WHERE p.campaign_id = (SELECT id FROM campaigns WHERE slug = $1)
ORDER BY p.total_engagements DESC
LIMIT 5;

2c. Campaign notes and direction

SELECT * FROM campaign_notes
WHERE campaign_id = (SELECT id FROM campaigns WHERE slug = $1)
ORDER BY source_date DESC
LIMIT 10;

Step 3: Build the Guardrails

Synthesize from corrections log and client direction into a DO/DON'T framework:

Language Guardrails

Pull from corrections.md terminology table and pillar-specific corrections:

• Use: [approved terms, phrases, positioning]
• Avoid: [flagged terms, phrases, approaches]
• Compliance notes: [legal requirements, disclaimers, character restrictions]

Creative Guardrails

Pull from corrections and client direction:

• Visual direction: [approved aesthetic, mood, style]
• Visual avoid: [flagged approaches — e.g., "faux pillars" for Day Well Spent]
• Format guidance: [recommended format based on data + client preference]

Tone Guardrails

Pull from pillar voice and Erica's golden rule:

• Voice for this pillar: [from campaigns.tone_voice or q2 goals]
• Erica's rule: "Clear and explicit, almost at the expense of cute copy"
• Pillar-specific tone notes: [from corrections/wins]

Step 4: Performance-Driven Recommendations

Based on what's worked historically:

What's resonated (from top performers)

• Identify common attributes of top-performing posts for this pillar:
  • Format (Carousel? Reel? Poll?)
  • Copy style (practical tips? questions? storytelling?)
  • Visual approach (people? graphics? text-heavy?)
  • Length/complexity

What to try next

• Based on engagement mix patterns from engagement-taxonomy.md:
  • If saves are high → lean into utility/reference content
  • If shares are high → lean into emotionally resonant content
  • If comments are high → lean into conversation-starting content
  • If engagement is low across the board → consider format change or platform shift

Platform-specific optimization

• Reference platform-behavior.md for the target platform:
  • Hook strategy (first 1-3 seconds for video, first slide for carousel)
  • Optimal length/slide count
  • Interactive elements (polls for Stories, CTA for Feed)
  • Algorithm signals to optimize for

Step 5: Generate the Brief

Output a structured content brief:

# Content Brief: [Pillar Name] — [Platform] — [Month Year]

## Strategic Context
**Pillar goal:** [from q2-2026-pillar-goals.md]
**This month's focus:** [specific angle or theme for this batch]
**Format:** [recommended format with rationale from performance data]
**Posts:** [count] for [platform]

## Messaging Direction
**Key message:** [one sentence — what should the audience take away?]
**Tone:** [voice descriptor for this pillar]
**Hook strategy:** [how to open — based on platform behavior]

## Content Specifications
**Format:** [Carousel/Reel/Static/Story/Short]
**Length:** [slide count, video duration, or copy length]
**Visual direction:** [mood, style, references to what's worked]
**CTA:** [desired audience action — save, share, comment, click]

## Guardrails

### DO
- [Approved language, approaches, formats — from wins + goals]
- [Specific things that have worked for this pillar]
- [Platform-specific best practices]

### DON'T
- [Flagged language — from corrections log]
- [Avoided approaches — from client feedback]
- [Platform anti-patterns]

### COMPLIANCE
- [Legal requirements — "terms apply," disclaimers]
- [Product name accuracy — from terminology table]
- [Brand standards — alt text requirements, logo usage]

## Performance Context
**Best performing post to date:**
- [Post summary, ER, key metrics]
- **Why it worked:** [behavioral analysis]

**What the data suggests for this batch:**
- [Data-driven recommendation]
- [Format/approach suggestion]

## Copy Framework
**Opening hook:** [suggested approach — question, stat, statement]
**Body:** [key points to cover, in priority order]
**Closing CTA:** [specific action + why]
**Hashtag strategy:** [if applicable — note: no keyword stuffing per corrections]
**Alt text:** [requirements — include "Synchrony logo" per corrections]

## Approval Notes
- [Who approves: Casey for strategy, Erica for language/compliance, legal for disclaimers]
- [Timeline: Tuesday submission → Wednesday review → 5-day legal SLA]
- [Any special approval requirements for this pillar]

Step 6: Validation Checklist

Before finalizing, run through:

• Does the messaging align with the pillar's Q2 strategic goal?
• Are all flagged terms from the corrections log avoided?
• Is the format recommendation supported by performance data?
• Is the tone consistent with Erica's golden rule?
• Are compliance requirements included?
• Is the approval workflow noted?
• Does the CTA match the desired engagement type for this pillar?

Important Rules

1. Corrections log is law. If the corrections log says don't use a term, don't include it in the brief — even if it would make the copy punchier.
2. Data drives format choice. Don't recommend a format just because it's trendy — recommend based on what's performed for THIS pillar on THIS platform.
3. Erica's golden rule is always relevant. "Clear and explicit, almost at the expense of cute copy."
4. Include the approval workflow. The creative team needs to know who approves and the timeline.
5. Brief ≠ final copy. The brief provides direction and guardrails — the creative team writes the actual content. Don't write the post for them unless asked.
6. One brief per pillar per platform. Don't combine platforms — each surface needs its own brief with platform-specific guidance.