e2e-audit
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands using
pnpmto run Playwright tests and start local development servers as defined in the scaffolding logic. - [DYNAMIC_EXECUTION]: The skill dynamically creates JavaScript/TypeScript test files (specifically
tests/demo/full-demo.spec.ts) based on discovered project routes and PRDs, and then executes these files using the Playwright runner. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from project PRDs and source code to guide its testing logic and report generation, which could be exploited if those files contain malicious instructions.
- Ingestion points: Project PRD files in
context-library/prds/and application source code/router files likesrc/App.tsx. - Boundary markers: No specific boundary markers or instructions to ignore embedded commands are identified when the agent reads these external context files.
- Capability inventory: The skill has the ability to execute shell commands (
pnpm) and perform file system writes (to.ralph/fix_plan.mdandoutputs/analyses/). - Sanitization: There is no evidence of sanitization, escaping, or validation of the content read from the PRDs or code before it is processed by the agent to generate reports or fix plans.
Audit Metadata