The Agent Skills Directory

[SAFE]: The skill instructions in SKILL.md define a clear and legitimate workflow for searching for camp providers using web search tools. The process relies on reading local configuration and profile files to personalize search results, which is consistent with its stated purpose.
[SAFE]: No obfuscation techniques such as Base64 encoding, zero-width characters, or homoglyphs were found in any of the skill files.
[SAFE]: The skill does not include any external dependencies, package installations, or remote script execution. It uses the agent's native capabilities for file system access and web browsing.
[SAFE]: There is no evidence of privilege escalation, persistence mechanisms, or metadata poisoning. The example files and templates serve as benign structural guides.
[SAFE]: While the skill exhibits an indirect prompt injection surface by processing untrusted data from web searches, this is an inherent risk of its functionality and no malicious patterns are present. Evidence chain: 1. Ingestion points: Web search results and local family-profile.md file. 2. Boundary markers: Not explicitly defined for delimiting external content. 3. Capability inventory: File creation and writing in the local research_dir. 4. Sanitization: Not explicitly defined for search-derived content.

Research Camps