v0.5.0: Add RBAC auth, LLM router, audit logging, output sanitization

- New: auth.py – Role-based access (operator/approver/viewer)
- New: llm_router.py – Configurable primary/fallback provider chain
- New: audit_log.py + audit_query.py – DSGVO-compliant audit trail
- New: sanitize_output() – masks secrets in bot responses
- Updated: bot.py – Integrated auth guards, router, audit logging
- Updated: ollama_client.py – Router-compatible interface
- Updated: CLOWNFISCHSERVER.md – Documentation updated
- Updated: README.md – Documentation updated

bot/shell.py

@@ -0,0 +1,85 @@
+#!/usr/bin/env python3
+"""
+🐠 Clownfischserver
+Author:  Mehlzoerwer-Claude (https://github.com/mehlzoerwer-claude)
+License: GPL-3.0 – Keep it open. Always.
+Repo:    https://github.com/mehlzoerwer-claude/clownfischserver
+"""
+
+import asyncio
+import logging
+import os
+import shlex
+
+logger = logging.getLogger(__name__)
+
+MGMT_USER = os.getenv("MGMT_USER", "clownfish")
+COMMAND_TIMEOUT = 120  # seconds
+
+
+async def run_shell(command: str) -> tuple[str, str, int]:
+    """Run a shell command as mgmt user, return (stdout, stderr, returncode)"""
+    try:
+        current_user = os.getenv("USER", "")
+        if current_user != MGMT_USER:
+            full_cmd = f"sudo -u {MGMT_USER} bash -c {shlex.quote(command)}"
+        else:
+            full_cmd = command
+
+        process = await asyncio.create_subprocess_shell(
+            full_cmd,
+            stdout=asyncio.subprocess.PIPE,
+            stderr=asyncio.subprocess.PIPE,
+            env={**os.environ, "HOME": f"/home/{MGMT_USER}"}
+        )
+
+        try:
+            stdout, stderr = await asyncio.wait_for(
+                process.communicate(),
+                timeout=COMMAND_TIMEOUT
+            )
+        except asyncio.TimeoutError:
+            process.kill()
+            return "", f"Timeout nach {COMMAND_TIMEOUT}s", 1
+
+        return (
+            stdout.decode("utf-8", errors="replace").strip(),
+            stderr.decode("utf-8", errors="replace").strip(),
+            process.returncode
+        )
+
+    except Exception as e:
+        logger.error(f"Shell execution error: {e}")
+        return "", str(e), 1
+
+
+async def execute_command(shell_cmd: str, ollama_client) -> str:
+    """
+    Execute a pre-determined shell command and return formatted output.
+    Optionally asks Ollama to explain errors or complex output.
+    """
+    logger.info(f"Executing: {shell_cmd}")
+
+    stdout, stderr, returncode = await run_shell(shell_cmd)
+
+    # Build response
+    response_parts = [f"$ {shell_cmd}"]
+
+    if stdout:
+        response_parts.append(stdout)
+
+    if stderr and returncode != 0:
+        response_parts.append(f"[STDERR]\n{stderr}")
+
+    if returncode != 0:
+        response_parts.append(f"[Exit code: {returncode}]")
+
+    raw_output = "\n".join(response_parts)
+
+    # If there's an error or output is complex, ask Ollama to explain
+    if returncode != 0 or len(stdout) > 500:
+        explanation = await ollama_client.explain_output(shell_cmd, stdout, stderr)
+        if explanation:
+            raw_output += f"\n\n💡 {explanation}"
+
+    return raw_output or "✓ Ausgeführt (keine Ausgabe)"