project-intelligence-timeline-dashboard
Pass
Audited by Gen Agent Trust Hub on Jul 23, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to Indirect Prompt Injection because it processes untrusted data and interpolates it into a generated HTML file without explicit sanitization or boundary enforcement.
- Ingestion points: Reads data from external sources including the
Change Events listandSHAREPOINT.md(Project Analysis Summary section). - Boundary markers: No explicit delimiters or instructions are provided to the agent to treat the ingested data as untrusted or to ignore any instructions embedded within that data.
- Capability inventory: The skill generates HTML/CSS content and writes files to the user's SharePoint environment.
- Sanitization: There are no instructions for the agent to perform HTML escaping, input validation, or content filtering on the project data (e.g., event titles, root cause statements) before rendering it into the final dashboard.
- [SAFE]: The skill instructions explicitly prohibit the use of external dependencies, CDN links, icon libraries, and external fonts, significantly reducing the supply chain and resource injection attack surface.
Audit Metadata