Phy Influencer Research

• Risk level: Low
• Analyzed: Mar 22, 2026, 2:01 PM
• Findings: 0
• Version: 1.0.2

Overview

The skill's instructions, inputs, and outputs are coherent with its stated purpose (automated influencer research); it is instruction-only, requests no credentials, and does not install software, though there are some operational ambiguities and user-facing cautions to consider.

This skill is coherent with its claimed purpose and does not request credentials or install code, but review these points before enabling it: - Autonomous runs: The skill supports an autonomous 'don't ask me' mode that will proceed with defaults (all platforms, 30–50 targets each) without confirmation. If you enable autonomous invocation, expect many web queries and file writes unless you constrain it. - Scraping and rate limits: The workflow instructs fetching data from multiple aggregator sites and running many parallel WebSearch calls. Verify this behavior complies with the target sites' terms of service and your organization’s web-scraping policies; consider adding rate limiting or caps to avoid IP blocks or legal/TOS issues. - Output location ambiguity: The example Python uses baseDir but baseDir is not defined in the SKILL.md. Decide where output files should be saved and whether the agent has permission to write there. - Data privacy: Influencer data may include personal information. Confirm how/where generated spreadsheets will be stored and who can access them. - Operational limits: If you want to avoid large workloads, specify platform and target_count explicitly when invoking (e.g., only TikTok, 20 targets), or require confirmation before executing. If these caveats are acceptable, the skill appears internally consistent; if not, ask the author/maintainer to clarify defaults, define baseDir, and add explicit confirmation and rate-limiting behavior before use.