Back to skill
Skillv1.0.0
ClawScan security
Service Business Prospecting · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignApr 28, 2026, 3:59 PM
- Verdict
- benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- The skill's requested actions and requirements align with its stated purpose (daily lead generation); it is instruction-only, requests no credentials or installs, and the runtime steps stay within the scope of web-based lead discovery and local file output.
- Guidance
- This skill appears internally consistent for lead generation, but before installing consider: (1) legal and policy compliance — scraping sites like Yelp/Facebook may violate terms of service; ensure your use and rate limits comply and you have permission to contact scraped parties; (2) privacy — the skill will collect personal contact info (owner names, emails, phone numbers); treat outputs as personal data and secure/storage/retention accordingly; (3) deliverability — verify the CSV path (<outputs-folder>) is correct and the agent has internet access; (4) quality controls — the SKILL.md prohibits fabrication but operational safeguards (validation, manual review) are advisable to avoid false leads; (5) monitoring — watch for IP blocking or CAPTCHAs from target sites and ensure the agent's behavior is throttled and identifiable to avoid abuse. If you need higher assurance about compliance or want to restrict data collection, ask the skill author for a privacy/compliance statement or add explicit rate-limiting and source whitelisting.
Review Dimensions
- Purpose & Capability
- okThe name/description (daily lead generation for local US service businesses) match the SKILL.md tasks: web search, contact enrichment, scoring, CSV output. The skill requests no unrelated credentials, binaries, or system paths.
- Instruction Scope
- okThe instructions are narrowly scoped to web discovery, enrichment (website/Facebook/Yelp), filtering, scoring, and saving CSV/notes locally. There are no directives to read unrelated system files, call hidden endpoints, or exfiltrate data to third parties. It does instruct collecting owner names, phones, and emails — which is appropriate for lead generation but is personal contact data.
- Install Mechanism
- okNo install spec and no code files are present (instruction-only). This is the lowest-risk install model: nothing is downloaded or written beyond the agent creating the described outputs.
- Credentials
- okNo environment variables, credentials, or config paths are required. The skill does not ask for unrelated secrets or cloud credentials.
- Persistence & Privilege
- okalways is false and the skill is user-invocable (normal). Model invocation is allowed (platform default) but the skill does not request elevated or persistent privileges or modify other skills.