Industry Research

This skill is plausible for industry/keyword research but contains several unexplained mismatches. Before installing or enabling it: - Confirm required binaries: verify that exa.js, firecrawl.js (and any MCP tooling) are intentionally available in the agent runtime. Ask the skill author to list required CLIs and versions. - Confirm credentials: the instructions reference AHREFS_API_KEY (and MCP access). Do not expose API keys unless you trust the skill and understand where outputs are written. Ask the author to declare required environment variables in the manifest. - Review file access: the skill will read .agents/product-marketing-context.md (and .claude/...) if present and will write .agents/industry-research-{client}.md. Ensure you’re comfortable with those reads/writes and that no sensitive files could be read inadvertently. - Scraping and legal risk: the workflow runs web-scraping tools and may hit third-party sites. Confirm scraping behavior complies with site terms and your organization’s policies. - Operational test: run it in a sandboxed agent or with fake/test credentials first to observe behavior (what network calls it makes, what files it reads/writes) before giving real credentials or enabling autonomous runs. If the author updates the manifest to declare the CLIs and AHREFS_API_KEY (or documents alternative no-credential modes) and documents exactly which files are read/written, the skill would be much easier to assess and less risky.