extension-builder
Pass
Audited by Gen Agent Trust Hub on Jul 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill's reference documentation (
references/autohand-extension-v1.md) references a JSON schema located on the official Autohand GitHub repository (github.com/autohandai). This is a well-known source and is used for standard configuration validation. - [COMMAND_EXECUTION]: The workflow involves using the
autohandCLI for extension validation and installation tasks. These are official vendor-provided tools belonging to the 'autohandai' ecosystem. - [PROMPT_INJECTION]: The skill implements specific security logic to mitigate indirect prompt injection. It instructs the agent to treat external Pi package sources as untrusted input, requiring it to process the text as data only and disregard any embedded prompts, workflow overrides, or credential requests found within the source code.
Audit Metadata