error-handling-recovery
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill instructions and reference patterns focus exclusively on best practices for error UX, accessibility, and graceful degradation. The logic is pedagogical and provides standard code snippets for validation and UI states without any malicious intent or hidden operations.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data by auditing the user's codebase for error-related strings and accepting user-provided support information for message generation. This represents an attack surface where malicious data in the codebase could influence agent behavior, though the scope is restricted to local code implementation. 1. Ingestion points: Local codebase audit via grep for error keywords and user-provided support details (SKILL.md, Step 0). 2. Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded commands within the scanned code. 3. Capability inventory: Modification of project files to implement error components and documentation (SKILL.md, Workflow Step 4). No network exfiltration or high-privilege command execution capabilities are utilized. 4. Sanitization: No explicit validation or sanitization is performed on the strings harvested from the codebase during the audit phase.
Audit Metadata