setup-devcontainer
Pass
Audited by Gen Agent Trust Hub on Aug 21, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches official CLI installers from
claude.aiandcursor.comto provision the development environment. - [REMOTE_CODE_EXECUTION]: Provisioning involves piping remote scripts from well-known service domains directly into a shell (
curl | bash). - [COMMAND_EXECUTION]: The skill generates and executes scripts for environment initialization, including user provisioning and identity file mounting.
- [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection (Category 8).
- Ingestion points: Reads repository configuration files like
package.json,go.mod, andCargo.tomlinSKILL.md(Step 1). - Boundary markers: None present to distinguish untrusted file content from instructions.
- Capability inventory: Generates and writes
Dockerfile,compose.override.yaml, and shell entrypoints. - Sanitization: No explicit sanitization or validation of the ingested repository metadata is performed before using it to generate executable files.
Audit Metadata