runbook-skill
Pass
Audited by Gen Agent Trust Hub on May 16, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions to download and install the
runbookCLI tool from the author's GitHub repository (Ariestar/agent-runbook). - [REMOTE_CODE_EXECUTION]: The setup process uses
cargo install --git, which downloads, builds, and executes source code from a remote repository on the local machine. - [COMMAND_EXECUTION]: The skill involves running shell commands, including
runbook scanfor environment discovery and potentiallycargo installfor initial setup. - [PROMPT_INJECTION]: The skill interprets output from the environment scan to derive an operating contract, which creates a surface for indirect prompt injection from repository metadata.
- Ingestion points: Command output from
runbook scanas described inSKILL.md. - Boundary markers: None identified to separate scanned data from agent instructions.
- Capability inventory: File searching (
rg), text replacement (sd), GitHub CLI (gh), and cloud/database CLI operations mentioned inSKILL.md. - Sanitization: No sanitization or validation of the scan results is specified before the agent uses them to form its execution plan.
Audit Metadata