Campaign Planner

Campaign Planner

Purpose

Strategic content calendar builder. Takes a brand's Business DNA, a campaign goal, and a timeframe — returns a structured posting schedule with weekly themes, content hooks, post types, and AI-suggested posting times. Feeds directly into Synthex's scheduling system for batch content creation and auto-publishing.

Workflow

1. Collect campaign parameters (see input collection below)
2. Load Business DNA for brand context
3. Generate calendar structure (weeks → themes → post slots)
4. Assign content types and hooks to each slot
5. Suggest optimal posting times per platform
6. Output as structured calendar + offer to generate content for each slot

Campaign Input Collection

Collect from user before planning:

CAMPAIGN PARAMETERS
  Goal:       [product launch / brand awareness / lead generation / seasonal / retention]
  Duration:   [30 / 60 / 90 days] — start date: [DD/MM/YYYY]
  Platforms:  [which platforms, or use all from Business DNA]
  Frequency:  [posts per week per platform — recommend 3-5 for most SMBs]
  Key dates:  [any launches, events, or deadlines within the period]
  Budget:     [organic only / includes paid ads]
  CTA:        [primary action — book / buy / DM / visit]

Calendar Structure

Phase Framework (30-day example)

For 60-day plans: repeat the cycle with a different campaign angle in weeks 5-8. For 90-day plans: add weeks 9-12 as retention/community phase.

Content Type Mix (per week)

Healthy content mix across all SMB accounts:

Platform Cadence (recommended for SMBs)

Times are suggestive — override with actual engagement data from Synthex analytics once available for the user's specific audience.

Calendar Output Format

Deliver as a structured week-by-week plan:

──────────────────────────────────────────
 CONTENT CALENDAR — [Brand Name]
 Period: [start date] → [end date]  |  Goal: [goal]
──────────────────────────────────────────

WEEK 1 — THEME: [theme name]
Focus: [week focus statement]

  MON [date]
    LinkedIn: Educational post — "3 ways [brand solves problem]"
    Instagram: Brand story — "[founder/team moment]"

  WED [date]
    LinkedIn: Case study hook — "[client result in numbers]"
    Instagram: Tip carousel — "[3-5 tips related to USP]"
    Facebook: Community question — "[discussion starter]"

  FRI [date]
    Instagram: Social proof — "[review or result]"
    TikTok: "[quick tip or behind-the-scenes script]"
    X: Thread — "[educational thread on core topic]"

WEEK 2 — THEME: [theme name]
...

──────────────────────────────────────────
CAMPAIGN SUMMARY
  Total posts: [count]  |  Platforms: [count]
  Content types: [breakdown]
  Key conversion dates: [list]
──────────────────────────────────────────

→ Generate content for all slots?
→ Or start with Week 1 only?

Batch Content Generation

After calendar approval, offer to generate content for each slot:

• Call brand-campaign-generator skill for each week as a batch
• Score all generated content via lib/ai/content-scorer.ts
• Flag any posts scoring below 75 for human review
• Pre-schedule approved posts via the scheduler API (app/api/scheduler/)

Visual assets for calendar slots are generated ONLY via the grounded pipeline — generateImage()/generateBatch() (lib/services/ai/image-generation.ts) or the generate_image/generate_video MCP tools, grounded-by-default on the owned reference library (public/reference-library/manifest.json). When planning visual-heavy slots, check library coverage for each subject first: subjects with no owned references will BLOCK generation — schedule real-photo capture/ingest (POST /api/admin/private-refs) ahead of those slots rather than falling back to stock or direct providers. Carpet-cleaning slots auto-apply the carpet-style-v1 LoRA (ccwcarpet); use the dashboard 3-variant batch + tap-to-rank to pick slot imagery.

Visual generation (binding): see .claude/rules/real-images-only.md + grounded-visuals. Direct provider calls fail CI.

Key Content Hooks by Goal

Product Launch:

• Week 1: Tease ("something's coming")
• Week 2: Reveal + benefits
• Week 3: Social proof + early access
• Week 4: Launch day + urgency + FOMO

Brand Awareness:

• Rotate: founder story → customer story → educational → behind the scenes

Lead Generation:

• Lead magnet → social proof → FAQ → offer → testimonial → offer

Seasonal:

• Pre-season (awareness) → peak (conversion) → post-season (retention)

Reference

• Business DNA: .claude/skills/business-dna/
• Content generation: .claude/skills/brand-campaign-generator/
• Platform adaptor: .claude/skills/platform-content-adaptor/
• Scheduling API: app/api/scheduler/
• Analytics (for timing optimisation): app/api/analytics/

Capability Uplift — Override Defaults

NEVER produce the default Mon/Wed/Fri three-pillar calendar (announce, educate, promote) applied generically across all brands. Never fill a calendar slot with "Educational post" without specifying what topic, what angle, and what platform-specific hook it uses.

INSTEAD every calendar slot specifies:

1. Platform (LinkedIn ≠ Instagram — different cadences)
2. Hook angle (the specific opening line concept, not just "educational")
3. Content pillar it belongs to (from Business DNA)
4. CTA direction (what action this piece drives)

Cadence varies by platform: LinkedIn 3×/week max, Instagram daily is fine, TikTok 1–2×/day is normal, Facebook 4×/week max, X/Twitter up to 3×/day.

REFERENCE .claude/skills/synthex-standards/references/content-standards.md

Foundation & Gate Wiring (SYN-1049)

Adopted from the senior-skill standard so every artefact this connector produces is checked against the locked foundation before it lands.

Reads at every invocation (never cached — re-read each run):

• .claude/memory/ceo-foundation.md — cadence map (Q2.5.3), universal + brand-specific taboos, frequency cap.
• .claude/memory/verification-gates.md — gate state for any claim referenced.

Output gate: every client-facing artefact this connector produces routes through brand-voice-enforce before the CEO batched-review queue. A REJECT blocks the artefact until the quoted offending string is fixed.

Evidence standard: every quantitative or factual claim carries exactly one tag — [VERIFIED] / [INFERENCE] / [UNCONFIRMED]. Untagged = defect (.claude/rules/fabel-evidence-standard.md). Never state a projected result as fact.

Spec: see spec.md in this skill directory.