fsbtactic-code/marketing-brain-skill / skills/marketing-brain
marketing-brain
FlaggedCreate, improve, research, or audit marketing strategy and execution using a bundled marketing library, verified live primary sources for volatile facts, and clearly labeled agent judgment. Use for full marketing systems, audits, landing pages and CRO, competitor intelligence, brand voice, audience and JTBD research, email sequences, SMM and social content systems, paid ads, headlines, copywriting, positioning, offers, launches, GTM, growth, funnels, retention, pricing, SEO, analytics, attribution, experiments, and consumer or buyer psychology. Trigger for English and Russian requests, including requests for ready-to-use assets.
references/workflow-email.md
View on GitHub# Email and Lifecycle Workflow
Use this workflow to design or write lifecycle email systems that are specific, truthful, measurable, and ready for implementation.
## Trigger conditions
- The user asks for welcome, onboarding, activation, nurture, launch, retention, renewal, win-back, or cart recovery email.
- The user asks for a newsletter system, lifecycle map, CRM sequence, triggered message, or email experiment.
- The user asks for outbound prospecting copy and provides a lawful audience and sending context.
- The user asks to audit or improve an existing email program.
Do not use this workflow for a single mechanical copy edit that introduces no new strategic or factual claim.
## Primary output
Create `EMAIL-SEQUENCES.md` when file output is requested or artifact mode is active. Otherwise return the same contract in the response.
The first line after the title must be `Workflow status: PASS`, `Workflow status: NEEDS_DATA`, or `Workflow status: BLOCKED`.
## Input contract
Required:
- Product, offer, real capabilities, current terms, and destination URL.
- Recipient segment, lifecycle state, entry event, and desired next state.
- One primary purpose and one primary CTA for each email.
- Market, language, sender identity, and brand voice or representative samples.
- Sending context, consent basis, suppression rules, and applicable jurisdiction.
Useful but optional:
- Audience research, JTBD, objections, interviews, support themes, and sales notes.
- Current flows, templates, performance exports, event taxonomy, and deliverability data.
- Baseline metrics with denominator, cohort, period, and attribution window.
- ESP, CRM, personalization fields, automation limits, and production constraints.
- Verified testimonials, cases, product screenshots, and approved legal language.
Ask only for missing inputs that change eligibility, claims, compliance, automation, or the requested asset.
## Dependencies
- Use user-provided product and customer data as the primary business context.
- Load `BRAND-VOICE.md` and `AUDIENCE-RESEARCH.md` when present and relevant.
- Run retrieval before selecting persuasion, sequence, or retention mechanics.
- Verify changing product terms, sender requirements, and legal rules from current primary sources.
- Never infer consent, customer status, proof, pricing, urgency, or scarcity.
## Retrieval plan
Run two or three focused queries adapted to the lifecycle state:
```text
python scripts/retrieve_marketing_context.py --query "lifecycle state customer job objection proof email sequence" --domain offer-messaging-copy --limit 5 --json
python scripts/retrieve_marketing_context.py --query "activation retention churn habit onboarding customer progress" --domain cro-funnels --limit 5 --json
python scripts/retrieve_marketing_context.py --query "email experiment primary metric guardrail consent deliverability" --domain measurement-analytics --limit 5 --include-discovery --json
```
For regulated, privacy-sensitive, or outbound work, also query `law-policy-ethics`. Open every discovery source before relying on it.
## Live official checks
- Verify current product price, offer, trial, cancellation, refund, and eligibility terms on canonical product pages.
- Verify the selected ESP or CRM fields, trigger behavior, suppression features, and message limits in its official documentation.
- Verify current sender authentication and bulk-sender requirements with the relevant mailbox provider documentation.
- Verify consent, identification, unsubscribe, recordkeeping, and privacy duties with official regulators for the target market.
- Verify links, forms, preference centers, unsubscribe paths, and destination claims before calling the flow implementation-ready.
- Record the verification date for every fact that can change.
Do not copy a universal open rate, click rate, send time, cadence, or deliverability rule from a legacy source.
## Execution phases
### Phase 1: Normalize the lifecycle brief
Define the recipient, current state, trigger, desired progress, business goal, offer, CTA, exclusions, and measurement horizon. Separate user facts, verified facts, assumptions, and open questions.
### Phase 2: Map the state transition
Describe why the recipient entered the flow, what they know, what they need next, what can block progress, and what event ends or changes the sequence. Include entry, exit, suppression, and conflict rules.
### Phase 3: Build the evidence and claim ledger
List every product claim, proof element, customer quote, deadline, price, and policy statement. Attach a source and observed date or remove it. Convert uncertain persuasion ideas into hypotheses.
### Phase 4: Design sequence architecture
Choose only the messages required for the state transition. For each message define its job, trigger or delay, audience rule, primary CTA, destination, and dependency. Do not force a fixed number of emails.
### Phase 5: Define message strategy
For each email specify recipient awareness, objection addressed, proof used, value delivered, voice, personalization, and reason for sending now. One email has one primary decision.
### Phase 6: Write production copy
Write the full subject, preview text, body, CTA, destination note, and required footer guidance. Use placeholders only for fields that the implementation system will populate and list their source fields.
### Phase 7: Specify automation and handoff
Define event names, delays, time zone behavior, frequency caps, suppression, branch conditions, retry behavior, ownership, QA seed list, and approval steps. Do not expose personal data in the artifact.
### Phase 8: Design measurement and experiments
Define delivery, engagement, state-transition, revenue, and complaint metrics with denominators and windows. For each experiment specify one variable, assignment unit, primary metric, guardrails, minimum detectable effect or decision threshold, and stop rule.
Do not prescribe a universal sample size, runtime, or expected lift. Use observed variance and volume or state what must be collected.
### Phase 9: Run preflight QA
Check rendering, links, tracking, personalization fallbacks, plain-text readability, accessibility, factual accuracy, consent, unsubscribe, suppression, mobile display, and consistency with the landing destination.
## EMAIL-SEQUENCES.md contract
Use these headings in this order:
1. Executive lifecycle decision
2. Workflow status and blocking data
3. Scope, audience, and assumptions
4. Product, offer, proof, and claim ledger
5. Lifecycle state map
6. Sequence architecture
7. Full email copy
8. Personalization and content rules
9. Automation, entry, exit, and suppression logic
10. Compliance and deliverability checks
11. Measurement plan
12. Experiment backlog
13. Implementation checklist
14. Evidence notes and open questions
For every email include `ID`, `Recipient state`, `Trigger or delay`, `Eligibility`, `Suppression`, `Message job`, `Subject`, `Preview text`, `Body`, `Primary CTA`, `Destination`, `Personalization fields`, `Proof source`, and `Success event`.
## Quality gate status
- `PASS`: all required inputs are known, claims and terms are verified, compliance and suppression logic are defined, full copy is complete, and no unresolved placeholder blocks implementation.
- `NEEDS_DATA`: a safe strategy or partial draft is possible, but product facts, consent, event data, proof, destination, automation fields, or measurement inputs are missing. List each missing item, owner, and affected section.
- `BLOCKED`: the requested flow depends on deception, fabricated proof, unlawful or unauthorized contact, hidden sender identity, prohibited targeting, unsafe regulated claims, or another condition that cannot be safely repaired inside the brief.
Assign exactly one status. Never label a flow `PASS` merely because copy exists.
## Stop and fallback rules
- Stop before full copy if the recipient, lifecycle state, product, or primary CTA is unknown.
- Stop before outbound copy if lawful contact and sender identification cannot be established.
- If brand voice is absent, state a minimal provisional voice and mark it `NEEDS_DATA` when approval is material.
- If proof is absent, write without proof claims and request approved evidence.
- If baseline metrics are absent, provide instrumentation and decision logic instead of forecasts.
- If live access is unavailable, mark volatile requirements unverified and do not call the flow launch-ready.
- If only one email is requested, keep the same evidence, compliance, and QA gates but omit irrelevant sequence sections.
## Completion criteria
The workflow is complete only when each message advances a defined customer state, all claims are supported, automation rules are executable, consent and suppression are addressed, measurement has valid denominators and decision rules, the assigned status is justified, and the user can hand the artifact to an implementation owner without guessing.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |