eve-new-project-setup
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions rely heavily on the execution of shell commands to configure the project environment, including CLI version checks, profile creation, authentication, and git operations.
- [EXTERNAL_DOWNLOADS]: The skill recommends the global installation of the
@eve-horizon/clipackage from the npm registry to enable the required functionality. - [INDIRECT_PROMPT_INJECTION]: The skill creates an attack surface by ingesting untrusted user input (Organization name, Project name, slugs, and Repository URL) and interpolating these values directly into CLI command arguments and YAML configuration files.
- Ingestion points: User input requested for organization/project metadata and git repository URLs in SKILL.md.
- Boundary markers: Absent; there are no explicit instructions for the agent to use delimiters or verify the integrity of the user-supplied strings before command interpolation.
- Capability inventory: Shell command execution (via
eve,git, andnpm) and file system modifications (writing to.eve/manifest.yamland creatingsecrets.env). - Sanitization: Absent; the skill does not include steps to sanitize or validate user input against shell injection or YAML malformation.
Audit Metadata