v1.1.0 to v1.1.0

1 added, 1 removed. Audit A to A.

---

name: ollama-localhost-guardian

- description: "Verifies local LLM services (e.g. Use for: ollama security, is ollama exposed, secure local LLM, ollama localhost."

+ description: "Verify that local LLM services, including Ollama, are bound and configured safely before they are exposed or connected. Use for: Ollama security, check exposed Ollama, secure local LLM, Ollama localhost."

version: 1.1.0

author: Stijnman

license: MIT

compatibility: Grok agent; optional MCP and shell access

metadata:

grok:

tags: [ollama security, is ollama exposed, secure local LLM, ollama localhost]

related_skills: [defensive-mcp-audit, exposed-service-triage, hitl-approver]

publication_reviewed: '2026-06-24'

---

# Ollama Localhost Guardian

## When to Use

- User says **ollama security** or task matches this capability

- User says **is ollama exposed** or task matches this capability

- User says **secure local LLM** or task matches this capability

- User says **ollama localhost** or task matches this capability

## Workflow

1. Check listener on default Ollama port (11434) via ss or netstat.

2. Pass if bound to 127.0.0.1; warn if 0.0.0.0 or ::.

3. Optionally check OLLAMA_HOST env and systemd override files (read-only).

4. Recommend binding to 127.0.0.1; hitl-approver before config changes.

5. Note installed AI packages (mcp, openai, etc.) for inventory only.

## Integrations

- `defensive-mcp-audit`

- `exposed-service-triage`

- `hitl-approver`

## Error Handling

| Failure | Response |

|---------|----------|

| Ollama not running | Report not listening; no install assumed. |

## Gotchas

- Check-only; config edits require user approval.

## Safety & Ethics (Publication-Ready)

This skill is designed for public distribution. Constraints:

- Read-only checks; no model pulling or inference API abuse.

- No exposure of API keys or model weights.

### Prohibited actions

- No unauthorized access, malware, or harmful automation

- No silent exfiltration of data, credentials, or telemetry

- No destructive system changes without hitl-approver

- No publication of user PII or environment secrets in outputs

## Example

**Input:** User request matching triggers above.

**Output:** Structured result per workflow; local artifacts only unless user opts in.