logic-locate

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [NO_CODE]: The skill consists entirely of Markdown instructions and guides. It does not include any scripts, executables, or configuration files that could be executed by the agent environment.
  • [SAFE]: No malicious patterns, such as prompt injection, obfuscation, or data exfiltration, were detected in the instructions or metadata. The skill follows standard analysis and reporting practices.
  • [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow to process untrusted user inputs, specifically stack traces and error messages, which represents a potential attack surface for indirect prompt injection.
  • Ingestion points: User-provided stack traces, error messages, and failing assertion logs described in SKILL.md and logic-locate-guide.md.
  • Boundary markers: The skill references external files (../_shared/common.md) for its 'Iron Law', which likely contains instructions for handling user content, though these are not present in the analyzed files.
  • Capability inventory: The skill is restricted to analytical reasoning and text-based reporting. There are no capabilities for shell command execution, file system modifications, or network communication.
  • Sanitization: No explicit sanitization or validation of the ingested stack traces is mentioned in the step-by-step logic guide.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 06:47 AM
Security Audit — agent-trust-hub — logic-locate