ceos-kickoff
Pass
Audited by Gen Agent Trust Hub on Mar 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a system command to synchronize the local repository with remote changes.
- Evidence: Uses the command
git -C <ceos_root> pull --ff-only --quietto update local data files from a remote source. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads and processes data from external markdown files which may contain instructions that influence agent behavior.
- Ingestion points: Reads content from files in directories such as
templates/anddata/(including vision, accountability, rocks, and scorecard data). - Boundary markers: No specific delimiters or instructions to ignore embedded commands within the data files were identified.
- Capability inventory: The skill has permissions to write files to the system and execute git commands.
- Sanitization: There is no evidence of content sanitization or validation performed on the read files prior to processing by the agent.
Audit Metadata