docker-init
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted local data from files like
package.jsonorREADME.mdto determine service dependencies. This is a necessary attack surface for the skill's primary function, and the potential impact is limited by the scoped tool access and the static nature of the generated templates. - Ingestion points: Project manifest files (e.g.,
package.json,pyproject.toml,Gemfile,go.mod,Cargo.toml), source code files,.env.example, andREADME.mdare read during the scanning phase described inSKILL.md. - Boundary markers: No specific boundary markers or 'ignore instructions' for the agent are used when reading the content of these files.
- Capability inventory: The skill utilizes file system tools (
Read,Write,Edit,Grep,Glob), user interaction (AskUserQuestion), and shell execution strictly limited to Docker commands (Bash(docker *)). - Sanitization: No explicit sanitization or validation of the content read from local files is performed before it is used to populate Docker configuration templates.
Audit Metadata