ollama-multi-model-writer

Pass

Audited by Gen Agent Trust Hub on Mar 18, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides a transparent local workflow for academic writing. It correctly uses local resources and follows its stated purpose without hidden behaviors.
  • [COMMAND_EXECUTION]: The skill uses a local script scripts/ollama-runner.sh to interface with the Ollama CLI. The script implementation uses proper shell quoting for variables, which prevents arbitrary command injection from user-provided inputs like {user_idea}.
  • [EXTERNAL_DOWNLOADS]: The documentation references model acquisition via the official Ollama registry, which is a well-known and trusted service.
  • [PROMPT_INJECTION]: The skill contains surfaces for indirect prompt injection. Ingestion points: {user_idea}, {results_data}, and {user_input} variables in SKILL.md. Boundary markers: double-quoting of arguments within the execution script. Capability inventory: local model execution via ollama CLI and writing results to markdown files. Sanitization: no explicit validation or escaping is applied to user-provided context before interpolation into model prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 18, 2026, 06:13 AM