play
Pass
Audited by Gen Agent Trust Hub on Nov 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external
.castrecording files provided via thefileargument. - Ingestion points: The agent reads and searches for
.castfiles using theGlobandBashtools. - Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the processed files.
- Capability inventory: The skill has access to
Bash,Glob, andAskUserQuestiontools. - Sanitization: No sanitization or validation of the content within the
.castfiles is specified before the agent acts upon them. - [DYNAMIC_EXECUTION]: The skill instructions include a 'Self-Evolving' section and 'Post-Execution Reflection' that explicitly direct the agent to modify the
SKILL.mdfile to fix issues or update parameters. - This self-modification mechanism allows for the dynamic alteration of the skill's logic at runtime, which could be leveraged to introduce malicious instructions if the agent's reasoning is influenced by adversarial input.
Audit Metadata