recording-browser-flow-as-test
Warn
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: MEDIUMDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [DYNAMIC_EXECUTION]: The skill generates Playwright test scripts (.spec.ts) based on snapshots of external web pages and recorded user interactions. These generated scripts are then executed locally via shell commands, posing a risk if the source content is crafted to inject malicious code into the test file.
- [INDIRECT_PROMPT_INJECTION]: The skill has a significant attack surface for indirect prompt injection as it ingests untrusted data from arbitrary web pages. Ingestion points: Accessibility tree data and DOM structures retrieved using browser_snapshot. Boundary markers: None identified; the skill does not specify delimiters to separate untrusted web content from system instructions. Capability inventory: Includes the ability to navigate URLs, interact with web elements, write files to the local repository, and execute shell commands. Sanitization: No evidence of sanitization or validation of the ingested web content before it is processed by the agent.
- [COMMAND_EXECUTION]: The skill requires the execution of shell commands (npx playwright test) to run the generated code, which effectively executes logic derived from untrusted external sources.
Audit Metadata