nextjs-developer
Warn
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: MEDIUMMETADATA_POISONINGINDIRECT_PROMPT_INJECTION
Full Analysis
- [METADATA_POISONING]: The skill metadata includes a non-existent model version ('gpt-5.4'). This is deceptive content intended to influence the host agent's perception of capabilities or manipulate model routing logic.
- [INDIRECT_PROMPT_INJECTION]: The skill operates by reading and modifying project code, which creates a vulnerability where instructions embedded in the code could influence the agent's actions.
- Ingestion points: Workspace files including Next.js components, route handlers, and server actions.
- Boundary markers: Absent. The skill lacks instructions to ignore or isolate natural language directives that may be hidden in code comments or string literals within the workspace.
- Capability inventory: The skill is configured with 'workspace-write' access, granting it the ability to modify files based on its analysis.
- Sanitization: Absent. There is no evidence of logic to sanitize or validate the content of the files it ingests before processing them.
Audit Metadata