setup-repo
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill includes a command to download and execute an installation script:
curl -fsSL https://raw.githubusercontent.com/holistics/anfra/main/install.sh | bash. This script is hosted on the vendor's official repository. - [EXTERNAL_DOWNLOADS]: The skill fetches a setup script from an external source (
raw.githubusercontent.com) to install the required CLI tool. - [CREDENTIALS_UNSAFE]: The skill manages sensitive database credentials, including passwords, private keys, and API secrets for multiple database providers. It mitigates security risks by instructing the agent to store these secrets exclusively in a local, git-ignored configuration file and explicitly warns against logging or committing credentials.
- [INDIRECT_PROMPT_INJECTION]: The skill processes repository files and external database schemas, presenting a potential surface for indirect instructions. Ingestion points: The
anfra ingestandanfra searchcommands read AML files from the repository. Boundary markers: The instructions do not specify delimiters or warnings for processed content. Capability inventory: The agent can execute CLI commands and initiate network connections to external databases. Sanitization: The skill focuses on structural validation rather than content sanitization.
Audit Metadata