drupal-canvas
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the user to download and execute the
@drupal-canvas/createscaffolding package via npx. It also recommends adding supplementary agent instructions fromdrupal-canvas/skillsusing thenpx skills addcommand. These resources are specific to the tool's ecosystem and appear to be legitimate developer utilities associated with the framework. - [COMMAND_EXECUTION]: Provides shell commands for project initialization and skill management using
npx. These commands are intended for environment setup and extending the agent's specialized knowledge for Drupal component development. - [INDIRECT_PROMPT_INJECTION]:
- Ingestion points: The skill defines components that interact with Drupal entities, fields, views, and APIs (referenced in
SKILL.mdunder Component Structure and Data Fetching). - Boundary markers: The example component logic and metadata templates do not include explicit instructions or delimiters to isolate data fetched from Drupal from the agent's instructional context.
- Capability inventory: The skill has the capability to execute shell commands (
npx) and install additional agent logic, which increases the potential impact if a downstream injection occurs. - Sanitization: The skill relies on standard React/TypeScript prop patterns for data rendering but lacks specific guidance for the agent to sanitize or validate external content to prevent it from being interpreted as instructions.
Audit Metadata