skill-builder
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes standard system utilities such as
go,rsync,perl,awk, andsedto manage skill packages and generate artifacts. These commands are used for legitimate build and transformation tasks. - [DYNAMIC_EXECUTION]: Through
scripts/run-ao.sh, the skill compiles its core logic from a local Go source directory and runs the binary. This self-contained build process is typical for this type of meta-tool. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external skill content during audits and conversions. It provides protection against indirect injection by following a 'clean-room' approach that excludes third-party prompts and prose from generated output, and it uses directory containment checks to prevent unauthorized file operations.
Audit Metadata