business-plan-generator
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [DYNAMIC_EXECUTION]: The skill instructions and workflow explicitly require the agent to generate a Python script based on a template and execute it to produce a real
.pptxfile. - [INDIRECT_PROMPT_INJECTION]:
- Ingestion points: The skill ingests untrusted user data in Step 1, where it asks the user for project names, descriptions, team backgrounds, and financial details.
- Boundary markers: The skill lacks explicit boundary markers or instructions for the agent to ignore potentially malicious commands embedded within the user's answers.
- Capability inventory: The workflow includes writing a script to a file and executing it, which allows for file system writes (saving the
.pptx) and command execution. - Sanitization: There are no instructions provided to sanitize, escape, or validate the user-provided strings before they are placed into the generated Python script's variable definitions. This creates a risk that a user could provide a string designed to break out of a Python string literal to execute arbitrary code when the script runs.
- [COMMAND_EXECUTION]: The skill requires the execution of the generated Python script via the shell to finalize the PowerPoint creation.
Audit Metadata