The Agent Skills Directory

[SAFE]: No prompt injection or safety bypass attempts detected. The instructions follow standard expertise-sharing patterns.
[SAFE]: No data exposure or exfiltration risks identified. The skill mentions reading local context files (e.g., .agents/product-marketing-context.md), which is a standard practice for personalizing agent responses.
[SAFE]: No obfuscation, hidden characters, or encoded payloads were found in any of the analyzed files.
[SAFE]: No remote code execution or unauthorized package installation patterns were detected. The skill mentions various CLI tools (stripe, posthog, etc.) for user reference rather than automated execution.
[SAFE]: No privilege escalation or persistence mechanisms identified.
[LOW]: Indirect prompt injection surface noted due to the skill's instruction to read external context files like product-marketing-context.md.
Ingestion points: .agents/product-marketing-context.md, .claude/product-marketing-context.md.
Boundary markers: Absent.
Capability inventory: The skill provides design advice and strategy; no active tool execution or file-write capabilities are defined within the skill itself.
Sanitization: Absent.

churn-prevention