cs-onboard
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill establishes a data flow where untrusted user input is captured and persisted, creating a surface for indirect prompt injection attacks.
- Ingestion points: The skill ingests data through user responses during the '7 Interview Dimensions' described in
SKILL.md. - Boundary markers: Data is stored in
~/.claude/company-context.mdusing a structured Markdown format (headers and bold keys). While this provides organizational structure, it does not prevent the interpretation of instructions embedded within the text values. - Capability inventory: The skill writes to the local file system (
~/.claude/company-context.md). This file is explicitly intended to be loaded and followed by other C-suite advisory skills (CEO, CFO, CTO advisors), expanding the impact of any injected instructions. - Sanitization: The instructions do not specify any programmatic sanitization or escaping of the user-provided text before it is written to the persistent context file, relying instead on the agent's adherence to privacy guidelines which can be bypassed via injection.
Audit Metadata