typo3-webmcp
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines an architecture for agents to interact with web page content, creating a potential surface for indirect prompt injection through data processed by the agent tools.
- Ingestion points: Agent tools ingest data from search results, news filters, article facts, and form fields on TYPO3 frontend pages as defined in the mapping instructions (SKILL.md, references/typo3-integration.md).
- Boundary markers: The skill mandates the use of
untrustedContentHintfor any output containing non-static text and requires visible human confirmation before any state-changing or consequential tool execution (SKILL.md, references/security-and-evidence.md). - Capability inventory: Capabilities include site search, navigation, factual page data retrieval, and form preparation/submission using existing, project-owned sitepackage endpoints (references/typo3-integration.md).
- Sanitization: The instructions strictly require robust server-side validation, CSRF protection, rate limiting, and script-safe JSON encoding to prevent malicious data from influencing the agent or the backend system (SKILL.md, references/security-and-evidence.md).
Audit Metadata