pickup

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly tells the agent to capture its parent process command line (including flags) with ps and reuse it verbatim to launch the successor, which will cause any API keys or secrets passed as command-line flags to be included in the generated launch command (exfiltration risk).

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.65). In Read mode, the required workflow reads a handoff document from ~/.agents/handoffs/<project>/ (which may have been authored by a different agent/user) and then “read[s] the chosen document in full” to brief the user—this free text is ingested into the LLM context, creating an indirect prompt-injection risk.