replit-ci-integration
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze and modify external GitHub repositories to configure CI/CD pipelines, creating a surface for indirect prompt injection.
- Ingestion points: The agent ingests untrusted data by reading file content, directory structures, and existing configuration from a user's GitHub repository via the
Readtool. - Boundary markers: The instructions do not define boundary markers or provide explicit prompts to the agent to ignore instructions embedded within the files it processes.
- Capability inventory: The skill allows the agent to use
Write,Edit, andBash(gh:*)tools, which could be exploited to inject malicious workflows or exfiltrate secrets if the agent follows instructions hidden in the repository content. - Sanitization: No sanitization or validation mechanisms are described to filter potentially malicious instructions from the ingested repository data before the agent performs write or configuration operations.
Audit Metadata