fnl-plugin-packaging
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONREMOTE_CODE_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill documents platform-specific CLI commands such as
claude plugin validate,agy plugin validate, andcodex plugin listfor manifest verification and plugin management. These are documented for legitimate development and validation workflows. - [EXTERNAL_DOWNLOADS]: References are made to official manifest schemas and documentation from trusted domains including
code.claude.com,developers.openai.com, andantigravity.google. These links point to official resources for the respective platforms. - [INDIRECT_PROMPT_INJECTION]: The distribution guidelines provide explicit instructions for agents to handle plugin installation by presenting lists to users and requiring manual confirmation, which serves as a mitigation against unauthorized automated actions.
- [REMOTE_CODE_EXECUTION]: The skill mentions installer scripts and tools like
npx skills addfor skill distribution. These methods are standard within the described development ecosystem and are directed towards user-initiated installation.
Audit Metadata