code-refactor
Pass
Audited by Gen Agent Trust Hub on Mar 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes the
jj(Jujutsu) command-line tool, specificallyjj diff -sandjj diff -- path, to inspect local file modifications. - [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface because it reads and processes external, untrusted source code that could contain adversarial instructions.
- Ingestion points: Untrusted data is ingested into the agent context through the output of
jj diffcommands which read local repository files. - Boundary markers: The prompt templates lack explicit delimiters or instructions to ignore embedded commands within the ingested code sections.
- Capability inventory: The agent possesses the capability to execute specific shell commands (
jj) and perform network-based information retrieval viaWebSearchandWebFetchtools. - Sanitization: No sanitization or validation of the ingested code content is performed before it is analyzed by the agent.
Audit Metadata