polaris-goal-compiler
Pass
Audited by Gen Agent Trust Hub on Oct 19, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists exclusively of markdown documentation and reference files. No malicious patterns, obfuscation, or unauthorized access attempts were detected within the provided files.
- [NO_CODE]: The skill does not include any executable scripts, binary files, or external software dependencies. It functions purely as a set of logical instructions for the AI agent to follow during task planning and execution.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a process for handling untrusted user goals (ingestion points in the Core Workflow). While there are no specific boundary markers or sanitization logic defined in the skill, it possesses no dangerous capabilities (no command execution, file system access, or network operations) that could be exploited by adversarial input. The protocol's focus on verification gates and claim ceilings acts as a defensive measure.
Audit Metadata