recoup-internal-consulting-linkedin-hooks
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided LinkedIn hooks for analysis and critique, which is a potential surface for indirect prompt injection if the input contains instructions aimed at diverting the agent.
- Ingestion points: User-supplied text to be graded or rewritten (e.g., "fix this hook").
- Boundary markers: The instructions provide a strict grading framework against specific patterns in
references/hook-patterns.md. - Capability inventory: The skill interacts with other internal consulting skills such as
recoup-internal-consulting-copy-writer. - Sanitization: None explicitly defined for the input strings.
- [COMMAND_EXECUTION]: The skill mentions a local maintenance script
swipe/_work/analyze_hooks.pyand a sync toolrecoup-internal-consulting-integrations-syncfor updating the reference dataset. - Evidence: Found in the "Keep it current" section of
SKILL.md. - Context: These references appear to be workflow instructions for maintaining the skill's performance data and target local project resources.
Audit Metadata