Outbound Prospecting Workflow
Analysis
This is a coherent instruction-only prospecting workflow that searches public web sources and drafts, but does not send, outreach emails.
Findings (2)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.
Use `web_search` to find the company's official website... Use `web_fetch` on key pages... `site:linkedin.com [Company Name] CEO`
The skill relies on external web search and fetching public pages, including searches for company personnel. This is disclosed and directly aligned with the prospecting purpose.
Checks for exposed credentials, poisoned memory or context, unclear communication boundaries, or sensitive data that could leave the user's control.
It follows the "Overseer Baseline" from MEMORY.md, operating in copilot mode only.
The skill points to stored instructions outside the visible SKILL.md. The visible text frames this as a copilot-only constraint, so this is a notice rather than a concern.
