plan-error-handling

Pass

Audited by Gen Agent Trust Hub on Oct 7, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted source code from a repository to identify error-handling gaps.
  • Ingestion points: The skill instructs the agent to 'Walk every error path' and 'Sweep' the codebase (SKILL.md), which involves reading arbitrary source files.
  • Boundary markers: There are no explicit instructions or delimiters defined to prevent the agent from following malicious instructions that might be embedded in the code comments or strings of the files being audited.
  • Capability inventory: While the skill is strictly 'plan-only' and forbids direct code edits, it generates a remediation plan that is intended to be executed by other skills (e.g., backend-observability). A successful injection could influence the generated plan to include malicious or weakened security configurations.
  • Sanitization: The skill does not specify any sanitization or filtering of the source code content before it is processed by the LLM.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 7, 2026, 11:26 PM