retrospect-report
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill passes raw user arguments directly to a bash script using the
$@shell expansion in SKILL.md. This creates a surface for argument injection, where a user could supply unintended flags or parameters that might alter the script's execution logic or target unauthorized files. - [INDIRECT_PROMPT_INJECTION]: The skill aggregates data from external session files and collaboration insights to generate reports. This ingestion of untrusted data constitutes an indirect prompt injection surface.
- Ingestion points: Reads session files and collaboration insights from the
.retro/directory. - Boundary markers: Absent. The instructions do not define delimiters or provide specific prompts to the agent to disregard instructions potentially embedded within the session data.
- Capability inventory: The skill utilizes
Bash,Read,Write, andGreptools, allowing for shell command execution and file system manipulation. - Sanitization: Absent. There is no evidence of sanitization or validation of the session data before it is computed into metrics or written into the final report.
Audit Metadata