content-script-developer
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides patterns for interacting with third-party web pages, which represents a potential attack surface where untrusted DOM content could influence agent logic.
- Ingestion points: Interactions with
document.bodyandMutationObserver(SKILL.md). - Boundary markers: Explicitly mentions that content scripts run in an isolated world and suggests keeping sensitive data out of the DOM.
- Capability inventory: Uses
chrome.runtime.sendMessagefor background communication anddocument.body.appendChildfor UI injection. - Sanitization: Recommends shadow roots and unique prefixes for styling to prevent host page pollution.
- [SAFE]: The provided code snippets follow legitimate extension development patterns and include performance safeguards like throttling. The author's provided website (shipshit.dev) is consistent with the skill metadata and represents a standard vendor resource.
Audit Metadata