generate-presentation

Pass

Audited by Gen Agent Trust Hub on May 11, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to its processing of untrusted data from URLs and local files.
  • Ingestion points: In SKILL.md (Step 1), the agent is instructed to fetch content from user-provided URLs using WebFetch and read local markdown files using the Read tool.
  • Boundary markers: The instructions lack boundary markers or specific guidance to the agent to treat fetched content as untrusted data, increasing the risk that embedded instructions could be followed.
  • Capability inventory: The skill has access to powerful tools including file writing (Write), shell command execution (Bash), browser automation (Playwright MCP), and a custom image generation tool (openai-gpt-image-mcp) that can write to absolute file paths.
  • Sanitization: No sanitization or validation of external content is performed before it is integrated into the prompt or used to drive subsequent agent steps.
  • [EXTERNAL_DOWNLOADS]: The skill frequently fetches content from external websites based on user input, which can be used to inject malicious content into the agent's workflow.
  • [COMMAND_EXECUTION]: The skill performs command-line operations, including running a bundled Python script for PDF conversion and utilizing a custom-built MCP server that handles file system writes to absolute paths provided as arguments.
Audit Metadata
Risk Level
SAFE
Analyzed
May 11, 2026, 10:49 AM
Security Audit — agent-trust-hub — generate-presentation