monorepo-navigator
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The script
scripts/impact_detector.pyuses thesubprocess.runfunction to executegit diffandgit merge-basecommands. This is used to identify which files have changed in the monorepo compared to a git reference (like origin/main). The commands are executed as lists (not through a shell), which is a secure practice, and the functionality is central to the skill's purpose of impact analysis. - [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it processes untrusted data from the repository being analyzed, such as package names, descriptions, and version information from
package.json,pnpm-workspace.yaml, andlerna.jsonfiles. * Ingestion points: The scriptsdependency_graph.py,impact_detector.py, andpackage_analyzer.pyall read and parse various workspace configuration files and command outputs from the monorepo root. * Boundary markers: There are no explicit instructions to the agent to ignore potentially malicious content embedded in these configuration files. * Capability inventory: The agent has the capability to run the provided Python scripts which execute git commands. * Sanitization: The scripts do not perform specific sanitization of the values read from configuration files, though they are parsed as JSON or through simple YAML logic.
Audit Metadata