wordpress-plugin-to-emdash
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's core workflow involves the agent inspecting and analyzing untrusted WordPress plugin source code, database schemas, and REST endpoints. This ingestion of third-party content creates a surface where malicious instructions embedded in the source code (e.g., within comments or string literals) could attempt to influence the agent's analysis or implementation logic.\n
- Ingestion points: WordPress plugin source code (PHP, JS, CSS), database configurations, and metadata provided in the source environment.\n
- Boundary markers: The instructions do not explicitly define delimiters or 'ignore-instructions' blocks for the agent to use while reading the untrusted source code.\n
- Capability inventory: The migration process results in the creation of code that can perform network fetches, access persistent storage (KV and collections), schedule background tasks, and manage user data.\n
- Sanitization: The risk is significantly mitigated by the skill's requirement to use sandboxed plugin formats by default, which enforces strict capability gating and explicit trust contracts for the generated implementation.
Audit Metadata