instagram-card-maker
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is designed to be self-contained and explicitly prohibits automated publishing, scheduling, or direct interaction with the Instagram API, which limits its security footprint to local asset generation with no risk of unauthorized account access or data exfiltration.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data through web research and user-supplied brand briefs, creating an attack surface for indirect prompt injection. Mandatory Evidence Chain:
- Ingestion points: Research step in SKILL.md utilizing web search and page fetch to gather facts.
- Boundary markers: Present; the skill instructs the agent to treat ingestible content as data rather than instructions and mandates grounding all claims in verifiable sources with accessed dates.
- Capability inventory: Web search, page fetch, file system write operations (to the .hyper/ workspace directory), and canvas-based image rendering.
- Sanitization: Present; the HTML composer in assets/card-studio.html uses textContent for rendering text and employs strict regex validation for base64-encoded image and font data URIs to prevent malicious payloads.
- [SAFE]: The HTML asset assets/card-studio.html follows security best practices by sanitizing user-supplied content and escaping HTML sensitive characters (e.g., replacing '<' with '\u003c') to prevent cross-site scripting (XSS) within the generated output files.
Audit Metadata