astrofy-sitespec-create
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized data access were detected. The skill follows a structured and validated workflow for task generation.
- [COMMAND_EXECUTION]: The skill executes the
astrofyCLI tool (astrofy sitespec list,astrofy skills list) to inspect project state and validate the generated task files. These commands are local to the developer environment. - [INDIRECT_PROMPT_INJECTION]: The skill processes external, potentially untrusted input from user requests (issues, messages, or code snippets) to derive checklist items, which is an inherent injection surface.
- Ingestion points: User-provided requests (referred to as
pedido) described inSKILL.md. - Boundary markers: The instructions do not specify explicit delimiters for the input data, but provide logic for clarifying ambiguities with the user before finalizing the task.
- Capability inventory: File writing to the local
.sitespec/directory and execution of theastrofyCLI utility. - Sanitization: The generated output is validated against a JSON schema via
astrofy sitespec list --jsonto ensure contract compliance.
Audit Metadata