The Agent Skills Directory

[SAFE]: The skill operates based on a set of 'Operating Principles' that enforce rigorous fact-checking and the use of a 4-tier source hierarchy to verify claims before presentation to the user.
[SAFE]: It implements 'Anti-Hallucination Rules' that explicitly forbid inventing data, review scores, or pricing, requiring the agent to search and verify all competitive intelligence.
[SAFE]: The skill includes instructions to flag all assumptions to the user, ensuring transparency and allowing for manual correction of potentially incorrect automated research.
[SAFE]: There are no signs of data exfiltration, unauthorized network operations, or hardcoded credentials. All network activity is restricted to the platform's standard web search and fetch tools for research purposes.
[SAFE]: Potential indirect prompt injection risks from reading untrusted external data (e.g., competitor reviews) are mitigated by mandatory source validation and the requirement to synthesize rather than directly output raw external content.

competitive-intelligence