Outbound Prospecting Workflow
v1.0.0Provides a structured workflow to research a target company, identify decision-makers, and draft personalized outreach emails using company-specific insights.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name and description match the runtime instructions: research a company, find decision-makers, and draft a personalized outreach email. The skill does not request unrelated credentials, binaries, or installs that would be out of scope for outbound prospecting.
Instruction Scope
Instructions are narrowly scoped to web_search and web_fetch of company pages and public LinkedIn searches, then drafting an email for human review. The SKILL.md references an external 'Overseer Baseline' (MEMORY.md) that is not included — this is a bookkeeping note rather than a direct instruction to access unrelated files. Also note: the skill's suggested LinkedIn scraping and extraction could expose personal data and may conflict with third-party Terms of Service or privacy rules; the skill does explicitly say the agent must not send the email itself.
Install Mechanism
This is an instruction-only skill with no install spec and no code files, so nothing will be written to disk or fetched during install.
Credentials
No environment variables, credentials, or config paths are requested. One caveat: the instructions assume the agent has web_search/web_fetch tools available — if those tools are backed by external services that require credentials, those credentials are not declared here and should be validated by the user/environment.
Persistence & Privilege
always is false and the skill does not request persistent system presence or modifications to other skills' configs. The skill allows model invocation (platform default), which is expected for this kind of helper and does not on its own raise concerns.
Assessment
This skill appears coherent and limited to researching companies and drafting emails for review. Before installing, confirm: (1) your agent's web_search/web_fetch tools are configured and you understand whether they send data to external services or require credentials; (2) you are comfortable with the privacy and Terms of Service implications of scraping LinkedIn and other public sites in your jurisdiction; (3) the agent is configured so it will not send outreach automatically — human review is needed; and (4) if the platform supplies a MEMORY.md or 'Overseer Baseline', review it to ensure no unexpected data access is implied. If any of those points are unclear, ask the skill author for clarification before use.Like a lobster shell, security has layers — review code before you run it.
latest
SKILL.md - Outbound Prospecting
This skill provides a structured workflow for researching and initiating contact with potential leads. It follows the "Overseer Baseline" from MEMORY.md, operating in copilot mode only.
Workflow
The process is designed to be run on a single target company domain or name.
1. Research and Audit (research_auditor)
- Objective: Gather intelligence on the target company to understand their business, identify potential needs, and find personalization points.
- Actions:
- Use
web_searchto find the company's official website, services, products, and recent news. - Use
web_fetchon key pages (Homepage, About Us, Blog) to extract detailed text. - Analyze the fetched content to identify:
- Core business offering.
- Target audience.
- Recent company announcements (product launches, funding, new hires).
- Potential pain points (e.g., outdated website, poor SEO, lack of recent blog posts).
- Use
2. Identify Decision-Maker (contact_finder)
- Objective: Find the most relevant person to contact at the company.
- Actions:
- Use
web_searchwith queries like:site:linkedin.com [Company Name] CEOsite:linkedin.com [Company Name] Foundersite:linkedin.com [Company Name] Head of Marketing
- Identify the name and title of a primary decision-maker.
- Use
3. Draft Outreach Email (draft_writer)
- Objective: Create a personalized, evidence-backed email ready for human review.
- Actions:
- Synthesize all gathered information.
- Draft an email that follows the hybrid tone: relationship-first opener + evidence-backed revenue-loss angle.
- The draft must include at least 3 company-specific facts discovered during the research phase.
- The final output should be a markdown block containing the draft, ready to be copied. The email itself should not be sent by the agent.
Comments
Loading comments...
