creative-iteration-playbook — Claude Code skill: what it does
creative-iteration-playbook turns social creative work into briefs, test matrices, a production pipeline and stop thresholds for paid campaigns.
Apache-2.0 ★ 450 updated 2026-04-03
npx skills add gtmagents/gtm-agents --skill creative-iteration-playbook -a claude-codeWhat creative-iteration-playbook does
Five stages hold the whole thing together: brief, test matrix, production pipeline, measurement and rollout rules. The brief covers hook, CTA, proof, edit guidance and the legal lines that must survive. The matrix spreads concept, format, CTA, length and hook across channels so a single test does not tangle them.
Templates ship with it: a one-slide brief and checklist, an experiment tracker with hypotheses and outcomes, an evergreen asset library layout with tag conventions, a per-channel hook and CTA bank, a runbook for keeping paid and organic teams in step (standup Monday, QA Wednesday, readout Friday), plus a guardrail sheet of minimum engagement and CVR per channel.
Other tools do part of the work. Serena patches landing pages, Playwright builds screenshot packs for QA, Context7 pulls platform specs, Sequential Thinking runs the retro. Success and kill thresholds go in before launch; the skill suggests writing them down first and gives CTR ≥1.5% for X and 35% hook retention on Reels as sample numbers. A paid boost pauses once CPA drifts more than 20% off baseline. Agree on the number first, or a failed campaign turns into an argument about taste.
What it helps with
- rewrite a brief for a fatigued campaign
- split one test into concept, format, CTA, length and hook variables
- log hypotheses and outcomes in a single experiment tracker
- set a kill threshold before a paid push goes live
- turn winning assets into reusable hook, format and CTA formulas
Example prompts for the agent
- Build a brief for the next Reels batch: hook, CTA, proof and legal notes.
- Map this test across concept, format and length, then open an experiment tracker.
- Run Playwright screenshots of the landing page for creative QA.
How to install creative-iteration-playbook
1Skills CLI (Vercel)
npx skills add gtmagents/gtm-agents --skill creative-iteration-playbook -a claude-code2Manually: put the files in place
~/.claude/skills/creative-iteration-playbook/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/creative-iteration-playbook/
1Skills CLI (Vercel)
npx skills add gtmagents/gtm-agents --skill creative-iteration-playbook -a codex2Manually: put the files in place
.agents/skills/creative-iteration-playbook/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add gtmagents/gtm-agents --skill creative-iteration-playbook -a cursor2Manually: put the files in place
.cursor/skills/creative-iteration-playbook/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install gtmagents/gtm-agents plugins/social-media-marketing/skills/creative-iteration-playbook2Manually: put the files in place
.github/skills/creative-iteration-playbook/SKILL.md1Skills CLI (Vercel)
npx skills add gtmagents/gtm-agents --skill creative-iteration-playbook -a gemini-cli2Platform CLI
gemini skills install https://github.com/gtmagents/gtm-agents --path plugins/social-media-marketing/skills/creative-iteration-playbookManually: put the files in place
.devin/skills/creative-iteration-playbook/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about creative-iteration-playbook
How do I install creative-iteration-playbook in Claude Code?
Run npx skills add gtmagents/gtm-agents --skill creative-iteration-playbook -a claude-code. The skill then sits at ~/.claude/skills/creative-iteration-playbook/SKILL.md.
What does creative-iteration-playbook do?
It takes creatives from brief to rollout: a channel test matrix, a production pipeline, UTM tagging with a scoreboard and insights log, and rules for scaling the winners.
Which tools does it work with?
Serena for landing page patches and Playwright for QA screenshot packs. Context7 brings platform specs, Sequential Thinking handles the post-test retro.
Source text
SKILL.md · Apache-2.0
--- name: creative-iteration-playbook description: Framework for briefing, testing, and scaling social creative variations. --- # Creative Iteration Playbook Skill ## When to Use - Refreshing creative fatigued campaigns. - Coordinating rapid creative experiments across paid + organic teams. - Documenting learnings with reusable creative formulas. ## Framework 1. **Creative Brief Template** – hook, CTA, proof, editing guidelines, mandatory/legal notes. 2. **Testing Matrix** – map variables (concept, format, CTA, length, hook) across channels. 3. **Production Workflow** – intake → concept → script/copy → edit → QA → publish. 4. **Measurement & Tagging** – UTM/tag schema, performance scoreboard, insights log. 5. **Scaling Guidelines** – roll-out thresholds, localization, evergreen library updates. ## Templates - Single-slide creative brief + checklist. - Experiment tracker w/ hypotheses, variants, outcomes, next steps. - Evergreen asset library structure with tagging conventions. - **GTM Agents Hook/CTA Library** – per-channel prompt bank pulled from GTM Agents Social Media Manager playbooks @puerto/README.md#183-212. - **Runbook for Paid + Organic Sync** – outlines Monday creative standup, Wednesday QA, Friday results readout. - **Guardrail Sheet** – min engagement/CVR per channel before pausing or pivoting. ## Tips - Pair qualitative notes (comments, watch time) with quantitative metrics. - Rotate creators and formats to avoid burnout and discover new winners. - Use with `build-social-calendar` and `campaign-optimization-manager` workflows. - Define success + kill thresholds before launching (e.g., CTR ≥1.5% for X, hook retention ≥35% on Reels). - Mirror GTM Agents escalation tree: Social Strategist → Marketing Director → Comms lead within 15 mins if sentiment spikes negative. - Keep Serena + Playwright handy for rapid landing page tweaks + visual QA. ## GTM Agents Creative Iteration Overlay 1. **Cadence** - Monday: backlog grooming + prioritization (use hook library to seed briefs). - Tuesday/Wednesday: production + QA (Pair design + copy pods, run Playwright visual checks). - Thursday: launch + monitor guardrails in near real-time. - Friday: learning sync → update evergreen library + share top performers in Weekly Packet. 2. **Guardrail Actions** - If two variants miss guardrails consecutively, trigger Sequential Thinking mini-RCA. - Pause paid boosts if CPA worsens >20% vs baseline; notify Growth + RevOps. 3. **Insight Logging** - Capture quote-sized "formula" (hook + format + CTA) for each winning asset; store with metrics + reuse plan. ## Tool Hooks - **Serena** – patch landing page snippets or marketing automation flows referenced in creative tests. - **Context7** – pull latest platform best-practice docs (LinkedIn ads specs, TikTok ratios) without leaving Claude. - **Sequential Thinking** – facilitate post-test retros and iteration hypotheses. - **Playwright** – automate screenshot packs for creative QA and stakeholder review. ---
© gtmagents · Apache-2.0 · gtmagents/gtm-agents