Openclaw Ai Marketing

Run, audit, or improve an OpenClaw/ClawLite-style AI marketing team as a coordinated multi-lane operating system. Use when the task is to operate or package...

Security Scan

An instruction-only skill that documents an AI-driven marketing operating model; its requirements and instructions are consistent with the stated purpose and it does not request secrets or install code.

Purpose & Capability

The name/description claim — running or packaging an OpenClaw/ClawLite-style marketing team — matches the provided content. The skill requires no binaries, env vars, or installs, which is appropriate for a playbook/operating-model skill.

Instruction Scope

SKILL.md is a runbook describing roles, receipts, and verification rules (e.g., real browser/public verification, URL/visibility proof, ASSET_CHECK). These instructions are consistent with operating a marketing team but are operationally broad: following them will typically require the agent to inspect live URLs, confirm publish status, and interact with external publishing systems. That network and verification activity is expected for the purpose but grants the agent broad discretion to perform external checks.

Install Mechanism

No install spec and no code files — lowest-risk delivery model. Nothing is written to disk and there are no remote downloads or packages to evaluate.

Credentials

No required environment variables, credentials, or config paths are declared. The open-source packaging guidance explicitly warns to strip private secrets and local paths, which aligns with the skill's nature.

Persistence & Privilege

always is false and the skill does not request persistent presence or modification of other skills/config. Autonomous invocation is allowed (platform default) but not combined with elevated privileges or credential access.

Assessment

This is a documentation/runbook skill (no code, no installs, no secrets requested) that instructs an agent how to operate and verify a multi-role AI marketing pipeline. It will expect the agent to check live URLs and produce verification receipts — so if the agent has web/browsing or publishing capabilities, it may perform external network checks and publish actions. Before installing, confirm you trust the agent's ability to perform external checks (and that those tools are configured safely), remove or avoid loading any private campaign data into the skill, and adjust aggressive defaults (e.g., the 12 blogs/day Tony target) to match your reality. If you plan to enable autonomous invocation for untrusted agents, consider the potential for unwanted publishing or external checks and restrict agent tooling accordingly.

This skill packages the ClawLite/OpenClaw AI marketing team into a reusable operating system.

What this skill is for

Use this skill when the goal is not “write one post” or “draft one blog.” Use it when the real task is to run or improve a daily AI marketing machine with handoffs, receipts, truth states, and conversion accountability.

This system assumes a multi-lane team:

• Hunter — community intel and X/Reddit learning
• JK — same-day content packaging / writing handoff
• Elon — social publishing
• Tony — blog publishing
• Jenny — lifecycle / activation email
• Peter — blog QA / closeout
• Karen — truth gate / QA
• Mission Control — state mirror and receipts

Core rule

The team is not complete when content exists. The team is complete only when:

• upstream research exists
• durable marketing-assets are updated
• downstream publishing is executed
• URLs / visibility / QA receipts exist
• Mission Control truth matches reality

System lanes

1. Hunter — upstream intelligence

Hunter owns:

• Reddit pain scan
• X pain scan
• X viral-learning scan
• Pain Map
• selection layer
• Intel Pack
• Mission Control mirror
• marketing-assets sync receipt

Hunter is incomplete if research stays trapped in notes.

2. JK — packaging layer

JK converts same-day inputs into cleaner content substrate for downstream lanes. JK should consume:

• same-day Hunter intel
• company positioning
• marketing-assets hooks / angles / proof / CTA

3. Elon — social lane

Elon owns publication, not just drafting. For each assigned lane/platform, require:

• same-day draft
• ASSET_CHECK
• post URL
• visibility proof
• acceptance receipt

X should preferentially run deep threads, not shallow one-liners.

4. Tony — blog lane

Tony is keyword-first and publish-first. Tony is not complete on “drafts.” Tony is complete only when the day’s blog target is actually published and verifiable.

Default rule in this system:

• Tony daily target = 12 blog publishes
• fewer than target must be called out as an explicit gap

5. Jenny — activation / lifecycle lane

Jenny owns:

• cohort selection
• send execution
• accepted send proof
• writeback/accounting
• same-day ASSET_CHECK

Do not confuse “send attempted” with “delivery complete.”

6. Peter — blog QA lane

Peter closes the Tony lane by verifying live/public reality. Peter PASS requires:

• correct live URL
• real browser/public verification
• clean QA receipt

7. Karen — truth gate

Karen does not create marketing. Karen verifies whether claimed completion is true.

8. Mission Control — truth mirror

Mission Control must reflect:

• current same-day truth
• partial completion vs real completion
• blockers without optimism inflation

Required operating principles

Asset-layer rule

Hunter learnings must be normalized into durable marketing-assets before the rest of the system can scale cleanly.

Receipt rule

Every lane leaves a dated receipt. No receipt = no completion.

ASSET_CHECK rule

All ClawLite/OpenClaw-facing growth/content lanes should leave ASSET_CHECK-backed evidence. Missing ASSET_CHECK means incomplete truth.

Truth-state rule

Use explicit states such as:

• DELIVERED
• PASS
• EXECUTED_BUT_BLOCKED
• BLOCKED_BUT_COMPLIANT
• AT_RISK
• SAME_DAY_UNRESOLVED_GAP

Do not collapse partial progress into fake completion.

Recommended execution order

1. Hunter research + X viral-learning
2. Pain Map + selection layer
3. marketing-assets sync
4. JK package if used
5. Elon social publishing
6. Tony blog publishing
7. Jenny activation / lifecycle
8. Peter closeout
9. Karen truth gate
10. Mission Control refresh

Open-source packaging guidance

When packaging this system as a reusable public skill:

• preserve the operating model, not private data
• keep role definitions, receipts, QA logic, and handoff discipline
• remove private secrets, private campaign data, and local-only credentials
• point users to configurable references rather than hard-coded company paths when needed

Read next when needed

• references/system-map.md
• references/role-contracts.md
• references/daily-loop.md
• references/open-source-packaging.md