demo-setup
Pass
Audited by Gen Agent Trust Hub on Oct 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local commands using
child_process.spawnSyncinscripts/check-setup.mjsto check for the presence and versions ofnode,ffmpeg,ffprobe, andpython3. This is used to ensure the environment is ready for video processing. - [EXTERNAL_DOWNLOADS]: During setup, the skill performs a vendoring step that downloads the recording logic from the author's repository (
AlexAnsart/demo-studio) using thenpx skills addutility. - [INDIRECT_PROMPT_INJECTION]: The skill analyzes local project files to automate configuration, which represents an ingestion point for untrusted data if the project files are malicious.
- Ingestion points: The skill reads
package.json,.env.example, anddemo.config.jsonfrom the root of the target project to infer stack settings and authentication requirements. - Boundary markers (absent): Ingested data is parsed directly as JSON or environment variables without specialized delimiters to prevent instruction injection.
- Capability inventory: The skill possesses capabilities for file system writes (
mkdirSync,appendFileSync), shell command execution viaspawnSync, and browser automation using Playwright. - Sanitization (present): Data is handled using standard programming parsers; the risk is low as the data is used to configure local templates and environment variables rather than being passed to external AI models.
Audit Metadata