grepai-embeddings-ollama

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPRIVILEGE_ESCALATIONPERSISTENCEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to download and execute an installation script from the official Ollama website (https://ollama.com/install.sh) via curl | sh. This is a documented installation path for a well-known service.
  • [COMMAND_EXECUTION]: Provides command-line examples for interacting with the Ollama binary, such as ollama serve, ollama pull, and ollama list. It also includes a rm command for clearing local index files during troubleshooting.
  • [PRIVILEGE_ESCALATION]: The documentation includes the use of sudo to manage Ollama as a system service via systemctl. These commands are standard for Linux service administration.
  • [PERSISTENCE]: Instructions include systemctl enable ollama, which configures the Ollama service to persist across system reboots.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to facilitate the indexing of codebases via GrepAI, creating a surface for processing potentially untrusted local data.
  • Ingestion points: The skill configures the environment for GrepAI to read local source code files (referenced in SKILL.md).
  • Boundary markers: No explicit prompt boundary markers or instruction-ignore warnings are present in the provided configuration snippets.
  • Capability inventory: The configured environment allows network requests via curl, file deletion via rm, and local model execution via ollama (found in SKILL.md).
  • Sanitization: Not applicable, as the skill provides static configuration instructions rather than active data processing logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 09:31 AM
Security Audit — agent-trust-hub — grepai-embeddings-ollama