agentic-ecology-init
Pass
Audited by Gen Agent Trust Hub on Oct 19, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches and installs related agent skills from established public registries (NPM) and official GitHub organizations using standard CLI tools.
- [COMMAND_EXECUTION]: Executes standard system and environment management commands such as
mkdir,uv, andnpxto scaffold the workspace, manage Python dependencies, and synchronize the virtual environment. - [REMOTE_CODE_EXECUTION]: Configures the project to pull a specific library directly from a repository hosted by a known research organization.
- [INDIRECT_PROMPT_INJECTION]:
- Ingestion points: Fetches package metadata from the NPM registry to provide valid skill name options to the user.
- Boundary markers: Data is processed via structured CLI tools (
jq) rather than direct prompt interpolation. - Capability inventory: Local file system writes within the project root and environment synchronization via
uv. - Sanitization: Registry data is filtered for specific keywords before being displayed to the user.
Audit Metadata