content-driven-development
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references the installation and usage of the
@adobe/aem-clipackage. This is a legitimate utility provided by the vendor for local development and previewing of Edge Delivery Services projects. - [COMMAND_EXECUTION]: The workflow utilizes standard command-line tools including
aem,npm,git, andgh(GitHub CLI) to manage the local development server, run linting/tests, and handle code deployment via pull requests. These are expected for the described development lifecycle. - [INDIRECT_PROMPT_INJECTION]: The skill ingests external content from URLs to validate block implementations against real-world scenarios. This is a core part of the 'Content Driven Development' philosophy described in the skill.
- Ingestion points: User-provided URLs in SKILL.md (Step 4) are fetched via
curlto verify availability and structure. - Boundary markers: No explicit instructions are provided to the agent to treat external content as untrusted strings, which is standard for this type of development utility where the developer reviews the output.
- Capability inventory: The skill utilizes shell commands (
aem,npm,git,gh,curl) as described in SKILL.md (Steps 1, 6, 8) and performs network requests to both local and Adobe-owned domains (e.g., da.live). - Sanitization: The skill relies on standard project tooling (linting and unit tests) for validation rather than specific prompt sanitization.
Audit Metadata