Add controlled Athena platform operator

This commit is contained in:
Mikei386
2026-08-23 21:10:44 +02:00
parent 2f4bff550e
commit 94f3758795
26 changed files with 1030 additions and 752 deletions
+1 -1
View File
@@ -64,7 +64,7 @@ else
fi
for optional in mike-ai-mcp-web mike-ai-mcp-homeassistant mike-ai-mcp-arr \
mike-ai-mcp-github mike-ai-mcp-platform-context mike-ai-mcp-athena-terminal \
mike-ai-mcp-github mike-ai-mcp-platform-context mike-ai-mcp-athena-operator \
mike-ai-mcp-unraid-official; do
if container_healthy "$optional"; then
pass "$optional aktiv"
+13
View File
@@ -0,0 +1,13 @@
FROM python:3.13-slim@sha256:ffb752e139c0a19692a43af8d8523b274222dd68eebad5d583b45c2201c6e30a
ARG MCP_PROXY_VERSION=0.12.0
ARG MCP_VERSION=1.29.0
RUN pip install --no-cache-dir "mcp==${MCP_VERSION}" "mcp-proxy==${MCP_PROXY_VERSION}" \
&& useradd --system --uid 10003 --create-home --home-dir /app operator
COPY athena_operator_mcp.py /app/athena_operator_mcp.py
RUN chown -R 10003:10003 /app
USER 10003:10003
WORKDIR /app
EXPOSE 8000
ENTRYPOINT ["mcp-proxy", "--host", "0.0.0.0", "--port", "8000", "--stateless", "--"]
CMD ["python", "/app/athena_operator_mcp.py"]
-20
View File
@@ -1,20 +0,0 @@
FROM python:3.13-slim@sha256:ffb752e139c0a19692a43af8d8523b274222dd68eebad5d583b45c2201c6e30a
ARG MCP_PROXY_VERSION=0.12.0
ARG MCP_VERSION=1.29.0
ARG PYYAML_VERSION=6.0.3
RUN apt-get update \
&& apt-get install -y --no-install-recommends bash file ripgrep \
&& rm -rf /var/lib/apt/lists/* \
&& pip install --no-cache-dir \
"mcp==${MCP_VERSION}" "mcp-proxy==${MCP_PROXY_VERSION}" "PyYAML==${PYYAML_VERSION}" \
&& useradd --system --uid 10002 --create-home --home-dir /app terminal
COPY athena_terminal_mcp.py /app/athena_terminal_mcp.py
RUN chown -R 10002:10002 /app
USER 10002:10002
WORKDIR /app
EXPOSE 8000
ENTRYPOINT ["mcp-proxy", "--host", "0.0.0.0", "--port", "8000", "--stateless", "--"]
CMD ["python", "/app/athena_terminal_mcp.py"]
+13 -11
View File
@@ -11,7 +11,7 @@ Prompts heraus, verhindert den früher beobachteten Kontextverbrauch von über
| Container | Endpunkt im Netz `mike-ai-tools` | Zweck | Standard |
|---|---|---|---|
| `mcp-platform-context` | `http://mike-ai-mcp-platform-context:8000/mcp` | Athena-Wissen, begrenzter Snapshot und kontrollierte Docs-Pflege | an |
| `mcp-athena-terminal` | `http://mike-ai-mcp-athena-terminal:8000/mcp` | fest begrenzte Lesebefehle und Syntaxprüfung für Athena-Quellen | an |
| `mcp-athena-operator` | `http://mike-ai-mcp-athena-operator:8000/mcp` | vollständiger Betrieb der Athena-KI-Plattform über Vorschau/Freigabe | an |
| `mcp-web` | `http://mike-ai-mcp-web:8000/mcp` | kompakte Websuche und Quellenvergleich | an |
| `mcp-homeassistant` | `http://mike-ai-mcp-homeassistant:8000/mcp` | Relay zum nativen HA-MCP; Token bleibt serverseitig | Profil `homeassistant` |
| `mcp-arr` | `http://mike-ai-mcp-arr:8000/mcp` | Sonarr/Radarr/Prowlarr mit serverseitiger Policy | Profil `arr` |
@@ -33,14 +33,16 @@ Host-Snapshot. Dokumentationspflege ist auf `docs/*.md` und einen zweistufigen
Preview/Approval-Ablauf begrenzt. Vollständige Beschreibung:
[`docs/PLATFORM_CONTEXT_MCP.md`](../../docs/PLATFORM_CONTEXT_MCP.md).
Der Athena Terminal MCP ist ebenfalls **keine allgemeine Shell**. Er führt nur
eine kurze serverseitige Positivliste von Lesewerkzeugen mit direkten
Argumenten aus (`ls`, `cat`, `head`, `tail`, `wc`, `stat`, `file`,
`sha256sum`, `grep`, `rg`, `sed`, `du`, `df`). Sichtbar sind ausschließlich
der read-only eingebundene versionierte Stack und der begrenzte
Laufzeitsnapshot. Shell-Syntax, Interpreter, Netzwerkprogramme, SSH/SCP,
Docker, `systemctl`, Prozesssignale, Reboot und Shutdown sind nicht vorhanden
beziehungsweise werden bereits vor der Ausführung abgewiesen.
Der Athena Operator MCP ist die einzige Bedienebene für Arbeiten an der lokalen
KI-Plattform. Qwen kann damit Quellen lesen, Änderungen vorbereiten, MCPs und
Docker-Dienste bauen/deployen, Modelle laden, Benchmarks starten, Profile und
OpenWebUI pflegen, Git veröffentlichen und Recovery erzeugen. Die unprivilegierte
MCP-Fassade sieht dabei nur einen lokalen Unix-Socket. Docker-Socket,
Repository, Modellverzeichnis, Git-Zugang und Root-Rechte verbleiben im
rootseitigen Executor. Jede Änderung benötigt eine vollständige Vorschau, ein
inhaltlich gebundenes, ablaufendes Ticket und eine spätere exakte Bestätigung.
Eine freie Shell sowie SSH-, Netzwerk-, Boot-, Kernel-, Treiber-, Partitions-,
Reboot- und Shutdown-Aktionen werden nicht angeboten.
TinySearch bleibt als Ganzes read-only. Nur das flüchtige tmpfs-Verzeichnis
`/home/tinysearch/.crawl4ai` ist beschreibbar, weil Crawl4AI dort seinen
@@ -77,7 +79,7 @@ passenden Server wählen:
| Serien, Filme, fehlende Episoden und Indexer-Releases | Sonarr und Radarr | Web |
| Persönliche Musikbibliothek, Titel, Alben, Künstler und Playlists | Navidrome | Web, ARR |
| Lesende NAS-, Docker-, Array-, Netzwerk- und Logdiagnose | Unraid (Systemdiagnose) | MUA |
| Athena-Quelldateien gezielt lesen oder Syntax prüfen | Athena Terminal | Platform Context für Architekturwissen |
| Athena-KI-Plattform entwickeln, testen, deployen, Modelle/Git/Recovery pflegen | Athena Operator | Platform Context für reine Architekturauskunft |
| Ausdrücklich benötigte MUA-Verwaltungsaktion | MUA | Unraid-Diagnose nicht parallel |
Ein leeres Ergebnis ist kein Grund, dieselbe Frage über mehrere unpassende
@@ -95,7 +97,7 @@ oder ein anderes Werkzeug benötigt wird.
`no-new-privileges`.
- Der SSH-basierte Unraid-Container ist nicht Teil des Standardstarts.
- Ein allgemeiner Host-Shell-MCP wird weiterhin bewusst nicht angeboten. Der
Athena Terminal MCP ist eine eng begrenzte Lesefassade, kein Host-Terminal.
Athena Operator besitzt strukturierte Plattformaktionen statt freier Befehle.
## Start
+174
View File
@@ -0,0 +1,174 @@
#!/usr/bin/env python3
"""Single MCP facade for end-to-end Athena platform operation."""
from __future__ import annotations
import json
import os
import socket
import sys
from typing import Any
VERSION = "1.0.0"
SOCKET_PATH = os.environ.get("ATHENA_OPERATOR_SOCKET", "/operator/operator.sock")
if hasattr(sys.stdin, "reconfigure"):
sys.stdin.reconfigure(encoding="utf-8", errors="replace")
if hasattr(sys.stdout, "reconfigure"):
sys.stdout.reconfigure(encoding="utf-8", errors="replace")
TOOLS = [
{
"name": "athena_operator_inspect",
"description": (
"USE FIRST for Athena administration. Inspect live platform state without changing it. "
"Subjects: overview, git_status, containers, models, jobs. This is the authoritative "
"starting point before MCP, Docker, model, profile, benchmark or recovery work."
),
"inputSchema": {
"type": "object",
"properties": {"subject": {"type": "string", "enum": ["overview", "git_status", "containers", "models", "jobs"], "default": "overview"}},
"additionalProperties": False,
},
},
{
"name": "athena_operator_read_source",
"description": (
"Read a versioned Athena repository file with SHA-256 and bounded line range. Use this "
"instead of guessing current Compose, MCP, router, model, OpenWebUI or recovery code."
),
"inputSchema": {
"type": "object",
"properties": {
"path": {"type": "string", "pattern": "^[A-Za-z0-9_.+/-]{1,240}$"},
"start_line": {"type": "integer", "minimum": 1, "maximum": 1000000, "default": 1},
"line_count": {"type": "integer", "minimum": 1, "maximum": 1000, "default": 300},
},
"required": ["path"], "additionalProperties": False,
},
},
{
"name": "athena_operator_search_source",
"description": "Search the complete versioned Athena repository for exact text before designing or modifying a component.",
"inputSchema": {"type": "object", "properties": {"query": {"type": "string", "minLength": 1, "maxLength": 200}}, "required": ["query"], "additionalProperties": False},
},
{
"name": "athena_operator_prepare",
"description": (
"PREPARE a state-changing Athena operation. This never changes state. It returns a "
"content-bound ticket, exact preview and confirmation phrase. Supported operations: "
"file_update (write repository and deployed stack files), run_checks, compose_deploy, "
"container_action, openwebui_sync, git_publish, model_download, benchmark, recovery. Show the complete "
"preview to the user and stop. Never execute in the same autonomous tool sequence. "
"file_update payload: {files:[{path,content,expected_sha256?}]}; run_checks: "
"{checks:[operator-tests,openwebui-filter-tests,platform-verify,compose-main,compose-mcp]}; "
"compose_deploy: {compose_file,services,build}; container_action: {action,containers}; "
"openwebui_sync: {}; "
"git_publish: {message}; model_download: {url,destination,sha256?}; benchmark: "
"{script,arguments}; recovery: {label}."
),
"inputSchema": {
"type": "object",
"properties": {
"operation": {"type": "string", "enum": ["file_update", "run_checks", "compose_deploy", "container_action", "openwebui_sync", "git_publish", "model_download", "benchmark", "recovery"]},
"payload": {"type": "object"},
},
"required": ["operation", "payload"], "additionalProperties": False,
},
},
{
"name": "athena_operator_execute",
"description": (
"EXECUTE exactly one previously prepared Athena operation. Call only after the user "
"approved the exact preview in a later message and supplied the exact confirmation. "
"Tickets expire, are content-bound and single-use. Long downloads, benchmarks and "
"recovery work return a job id. This tool cannot alter SSH, networking, WireGuard, "
"firewall, boot, kernel, drivers, partitions, mounts, shutdown or reboot."
),
"inputSchema": {
"type": "object",
"properties": {
"ticket": {"type": "string", "pattern": "^[0-9a-f]{32}$"},
"confirmation": {"type": "string", "pattern": "^EXECUTE [0-9a-f]{32}$"},
},
"required": ["ticket", "confirmation"], "additionalProperties": False,
},
},
{
"name": "athena_operator_job",
"description": "Poll one asynchronous model download, benchmark or recovery job. Do not start duplicate jobs while it is running.",
"inputSchema": {"type": "object", "properties": {"job_id": {"type": "string", "pattern": "^[0-9a-f]{32}$"}}, "required": ["job_id"], "additionalProperties": False},
},
]
def request(action: str, arguments: dict[str, Any]) -> dict[str, Any]:
payload = json.dumps({"action": action, "arguments": arguments}, ensure_ascii=False, separators=(",", ":")).encode() + b"\n"
with socket.socket(socket.AF_UNIX, socket.SOCK_STREAM) as client:
client.settimeout(20)
client.connect(SOCKET_PATH)
client.sendall(payload)
chunks = bytearray()
while not chunks.endswith(b"\n"):
part = client.recv(65536)
if not part:
break
chunks.extend(part)
if len(chunks) > 2_000_000:
raise RuntimeError("operator response exceeds limit")
response = json.loads(chunks)
if not response.get("ok"):
raise RuntimeError(response.get("error", "operator request failed"))
return response["result"]
def call(name: str, arguments: dict[str, Any]) -> dict[str, Any]:
mapping = {
"athena_operator_inspect": "inspect",
"athena_operator_read_source": "read_source",
"athena_operator_search_source": "search_source",
"athena_operator_prepare": "prepare",
"athena_operator_execute": "execute",
"athena_operator_job": "job",
}
if name not in mapping:
raise ValueError("unknown tool")
return request(mapping[name], arguments)
def emit(request_id: Any, result: Any = None, error: dict[str, Any] | None = None) -> None:
message = {"jsonrpc": "2.0", "id": request_id}
message["error" if error else "result"] = error or result
sys.stdout.write(json.dumps(message, ensure_ascii=False, separators=(",", ":")) + "\n")
sys.stdout.flush()
def handle(message: dict[str, Any]) -> None:
method, request_id = message.get("method"), message.get("id")
if method == "initialize":
emit(request_id, {"protocolVersion": message.get("params", {}).get("protocolVersion", "2024-11-05"), "capabilities": {"tools": {"listChanged": False}}, "serverInfo": {"name": "mike-ai-athena-operator", "version": VERSION}})
elif method == "tools/list":
emit(request_id, {"tools": TOOLS})
elif method == "tools/call":
params = message.get("params", {})
try:
value = call(str(params.get("name", "")), params.get("arguments") or {})
emit(request_id, {"content": [{"type": "text", "text": json.dumps(value, ensure_ascii=False, separators=(",", ":"))}], "structuredContent": value, "isError": False})
except Exception as exc:
emit(request_id, {"content": [{"type": "text", "text": f"ERROR: {exc}"}], "isError": True})
elif request_id is not None:
emit(request_id, error={"code": -32601, "message": "method not found"})
def main() -> None:
for line in sys.stdin:
try:
if line.strip(): handle(json.loads(line))
except Exception as exc:
sys.stderr.write(f"MCP input error: {exc}\n"); sys.stderr.flush()
if __name__ == "__main__":
main()
-547
View File
@@ -1,547 +0,0 @@
#!/usr/bin/env python3
"""Capability-bounded terminal MCP for the Athena source tree.
This is deliberately not a general shell. Commands are executed without a
shell, against read-only mounts, with a fixed environment and a strict program
and argument allowlist. The container has no Docker socket, host PID namespace,
SSH material, secrets or egress network.
"""
from __future__ import annotations
import ast
import json
import os
import re
import subprocess
import sys
from pathlib import Path
from typing import Any
SERVER_VERSION = "1.0.0"
WORKSPACE_ROOT = Path(os.environ.get("ATHENA_TERMINAL_WORKSPACE", "/workspace")).resolve()
RUNTIME_ROOT = Path(os.environ.get("ATHENA_TERMINAL_RUNTIME", "/runtime")).resolve()
MAX_OUTPUT_CHARS = 20_000
MAX_ARGUMENTS = 32
COMMAND_TIMEOUT_SECONDS = 8
ALLOWED_ROOTS = (WORKSPACE_ROOT, RUNTIME_ROOT)
BLOCKED_PROGRAMS = {
"ash", "bash", "busybox", "chroot", "curl", "dash", "docker", "env",
"fish", "ftp", "halt", "init", "kill", "killall", "ksh", "mount",
"nc", "netcat", "nft", "nohup", "perl", "php", "pkill", "poweroff",
"python", "python3", "reboot", "rsync", "scp", "sh", "shutdown", "ssh",
"sftp", "socat", "sudo", "su", "systemctl", "telnet", "umount", "wget",
"zsh",
}
ALLOWED_PROGRAMS = {
"cat", "df", "du", "file", "grep", "head", "ls", "rg", "sed",
"sha256sum", "stat", "tail", "wc",
}
SENSITIVE_PATH_TOKENS = {
".env", "authorized_keys", "agekey", "credentials", "id_ed25519",
"id_rsa", "private_key", "secret", "secrets", "shadow",
}
if hasattr(sys.stdin, "reconfigure"):
sys.stdin.reconfigure(encoding="utf-8", errors="replace")
if hasattr(sys.stdout, "reconfigure"):
sys.stdout.reconfigure(encoding="utf-8", errors="replace")
TOOLS = [
{
"name": "athena_terminal_policy",
"description": (
"USE FIRST before terminal work on Athena. Returns the exact capability boundary, "
"allowed read-only programs, visible roots and explicitly unavailable operations. "
"This tool performs no command. The terminal is not a shell and cannot access SSH, "
"Docker control, host services, secrets, networking, shutdown or reboot."
),
"inputSchema": {"type": "object", "properties": {}, "additionalProperties": False},
},
{
"name": "athena_terminal_run",
"description": (
"Run one bounded read-only command against Athena's versioned stack or bounded "
"runtime snapshot. Pass a program and an argument array; shell syntax, pipelines, "
"redirection and command substitution do not exist. Allowed programs are ls, cat, "
"head, tail, wc, stat, file, sha256sum, grep, rg, sed, du and df. Paths may only be "
"inside /workspace or /runtime. Never claim this can deploy, edit, control Docker, "
"restart services, use SSH or alter the host."
),
"inputSchema": {
"type": "object",
"properties": {
"program": {"type": "string", "enum": sorted(ALLOWED_PROGRAMS)},
"arguments": {
"type": "array",
"maxItems": MAX_ARGUMENTS,
"items": {"type": "string", "maxLength": 500},
"default": [],
},
"working_directory": {
"type": "string",
"enum": ["workspace", "runtime"],
"default": "workspace",
},
},
"required": ["program"],
"additionalProperties": False,
},
},
{
"name": "athena_terminal_validate_source",
"description": (
"Validate exactly one versioned source file without executing it. Supports Python "
"AST parsing, Bash syntax-only parsing, JSON parsing and YAML parsing. The file must "
"be below /workspace and cannot be a secret-bearing path. This does not build, run, "
"deploy or modify anything."
),
"inputSchema": {
"type": "object",
"properties": {
"path": {
"type": "string",
"minLength": 1,
"maxLength": 240,
"pattern": "^[A-Za-z0-9_./-]+$",
},
"kind": {
"type": "string",
"enum": ["auto", "python", "shell", "json", "yaml"],
"default": "auto",
},
},
"required": ["path"],
"additionalProperties": False,
},
},
]
def json_text(value: Any) -> str:
return json.dumps(value, ensure_ascii=False, separators=(",", ":"))
def within_root(path: Path, root: Path) -> bool:
try:
path.relative_to(root)
return True
except ValueError:
return False
def reject_sensitive_path(path: Path) -> None:
lowered_parts = {part.casefold() for part in path.parts}
lowered_name = path.name.casefold()
if lowered_parts & SENSITIVE_PATH_TOKENS:
raise PermissionError("secret-bearing paths are not accessible")
if any(token in lowered_name for token in ("secret", "credential", "agekey", "private")):
raise PermissionError("secret-bearing paths are not accessible")
def safe_path(value: str, cwd: Path, *, must_exist: bool = True) -> Path:
if not value or "\x00" in value or "\n" in value or "\r" in value:
raise ValueError("invalid path")
candidate = Path(value)
candidate = candidate if candidate.is_absolute() else cwd / candidate
resolved = candidate.resolve(strict=False)
if not any(within_root(resolved, root) for root in ALLOWED_ROOTS):
raise PermissionError("path is outside the allowed terminal roots")
reject_sensitive_path(resolved)
if must_exist and not resolved.exists():
raise FileNotFoundError("path does not exist")
return resolved
def clean_scalar(value: str) -> str:
if not isinstance(value, str) or len(value) > 500:
raise ValueError("invalid argument")
if any(char in value for char in ("\x00", "\n", "\r")):
raise ValueError("multiline and NUL arguments are forbidden")
if any(token in value for token in ("$(", "${", "`", ">", "<", "|", ";", "&&", "||")):
raise ValueError("shell syntax is forbidden")
return value
def path_argument(value: str, cwd: Path) -> str:
if value == "-":
raise ValueError("stdin paths are not supported")
return str(safe_path(value, cwd))
def validate_arguments(program: str, arguments: list[str], cwd: Path) -> list[str]:
if program in BLOCKED_PROGRAMS or program not in ALLOWED_PROGRAMS:
raise PermissionError("program is not allowed")
if len(arguments) > MAX_ARGUMENTS:
raise ValueError("too many arguments")
args = [clean_scalar(value) for value in arguments]
if program == "cat":
if not args:
raise ValueError("cat requires at least one file")
return [path_argument(value, cwd) for value in args]
if program in {"sha256sum", "file"}:
allowed_flags = {"-b"} if program == "file" else set()
result = []
for value in args:
if value.startswith("-"):
if value not in allowed_flags:
raise ValueError("unsupported option")
result.append(value)
else:
result.append(path_argument(value, cwd))
if not any(not value.startswith("-") for value in args):
raise ValueError(f"{program} requires a path")
return result
if program in {"head", "tail"}:
result = []
index = 0
if len(args) >= 2 and args[0] == "-n":
if not re.fullmatch(r"[0-9]{1,5}", args[1]) or int(args[1]) > 2000:
raise ValueError("line count must be between 0 and 2000")
result.extend(args[:2])
index = 2
paths = args[index:]
if not paths:
raise ValueError(f"{program} requires a path")
result.extend(path_argument(value, cwd) for value in paths)
return result
if program == "wc":
allowed = {"-c", "-l", "-m", "-w"}
result = []
paths = 0
for value in args:
if value.startswith("-"):
if value not in allowed:
raise ValueError("unsupported wc option")
result.append(value)
else:
result.append(path_argument(value, cwd))
paths += 1
if paths == 0:
raise ValueError("wc requires a path")
return result
if program == "stat":
if not args:
raise ValueError("stat requires a path")
if any(value.startswith("-") for value in args):
raise ValueError("stat options are not supported")
return [path_argument(value, cwd) for value in args]
if program == "ls":
allowed = {"-1", "-a", "-al", "-d", "-h", "-l", "-la", "-lh", "-R"}
result = []
for value in args:
if value.startswith("-"):
if value not in allowed:
raise ValueError("unsupported ls option")
result.append(value)
else:
result.append(path_argument(value, cwd))
if not any(not value.startswith("-") for value in args):
result.append(str(cwd))
return result
if program == "sed":
if len(args) < 3 or args[0] != "-n":
raise ValueError("sed only supports: -n START[,END]p FILE...")
if not re.fullmatch(r"[0-9]{1,7}(,[0-9]{1,7})?p", args[1]):
raise ValueError("sed expression is limited to printing a line range")
return ["-n", args[1], *[path_argument(value, cwd) for value in args[2:]]]
if program == "grep":
# GNU grep -R follows symlinks. Only -r is permitted because it skips
# directory symlinks and therefore preserves the visible-root boundary.
allowed = {"-c", "-E", "-F", "-i", "-l", "-n", "-r"}
result = []
index = 0
while index < len(args) and args[index].startswith("-"):
if args[index] not in allowed:
raise ValueError("unsupported grep option")
result.append(args[index])
index += 1
if index >= len(args):
raise ValueError("grep requires a pattern")
result.append(args[index])
index += 1
paths = args[index:] or [str(cwd)]
result.extend(path_argument(value, cwd) for value in paths)
return result
if program == "rg":
allowed_flags = {"--files", "--hidden", "--json", "-c", "-F", "-i", "-l", "-n"}
result = []
index = 0
files_mode = False
while index < len(args) and args[index].startswith("-"):
value = args[index]
if value in {"-g", "--glob"}:
if index + 1 >= len(args):
raise ValueError("missing glob value")
result.extend([value, args[index + 1]])
index += 2
continue
if value not in allowed_flags:
raise ValueError("unsupported rg option")
files_mode = files_mode or value == "--files"
result.append(value)
index += 1
if not files_mode:
if index >= len(args):
raise ValueError("rg requires a pattern")
result.append(args[index])
index += 1
paths = args[index:] or [str(cwd)]
result.extend(path_argument(value, cwd) for value in paths)
return result
if program == "du":
allowed = {"-a", "-h", "-s", "-sh"}
result = []
for value in args:
if value.startswith("--max-depth="):
depth = value.split("=", 1)[1]
if not depth.isdigit() or int(depth) > 5:
raise ValueError("du max depth must be between 0 and 5")
result.append(value)
elif value.startswith("-"):
if value not in allowed:
raise ValueError("unsupported du option")
result.append(value)
else:
result.append(path_argument(value, cwd))
if not any(not value.startswith("-") for value in args):
result.append(str(cwd))
return result
if program == "df":
allowed = {"-h", "-T", "-hT", "-Th"}
result = []
for value in args:
if value.startswith("-"):
if value not in allowed:
raise ValueError("unsupported df option")
result.append(value)
else:
result.append(path_argument(value, cwd))
return result
raise PermissionError("program policy is incomplete")
def compact_output(text: str) -> tuple[str, bool]:
if len(text) <= MAX_OUTPUT_CHARS:
return text, False
marker = f"\n...[output truncated from {len(text)} characters]...\n"
remaining = MAX_OUTPUT_CHARS - len(marker)
return text[: int(remaining * 0.75)] + marker + text[-int(remaining * 0.25) :], True
def policy() -> dict[str, Any]:
return {
"mode": "bounded-read-only-terminal",
"allowed_programs": sorted(ALLOWED_PROGRAMS),
"visible_roots": [str(root) for root in ALLOWED_ROOTS],
"execution": "direct argv only; no shell, pipelines, redirection or substitution",
"limits": {
"timeout_seconds": COMMAND_TIMEOUT_SECONDS,
"max_arguments": MAX_ARGUMENTS,
"max_output_characters": MAX_OUTPUT_CHARS,
},
"unavailable": [
"SSH/SCP/SFTP and all remote login",
"shutdown, reboot, halt and power operations",
"Docker socket, Docker control and container exec",
"systemctl, service control, process signals and host PID namespace",
"network clients, internet access, VPN/firewall/routing changes",
"interpreters, arbitrary scripts and package installation",
"writes to the Athena stack, host filesystem, Git or secrets",
],
"instruction": (
"This terminal supplies evidence and syntax validation only. Use a separate, "
"ticket-bound operator workflow for future deployments or state changes."
),
}
def run_command(arguments: dict[str, Any]) -> dict[str, Any]:
program = str(arguments.get("program", ""))
raw_args = arguments.get("arguments") or []
if not isinstance(raw_args, list) or not all(isinstance(value, str) for value in raw_args):
raise ValueError("arguments must be a string array")
cwd_name = str(arguments.get("working_directory", "workspace"))
cwd = WORKSPACE_ROOT if cwd_name == "workspace" else RUNTIME_ROOT if cwd_name == "runtime" else None
if cwd is None or not cwd.is_dir():
raise ValueError("working directory is unavailable")
argv = [program, *validate_arguments(program, raw_args, cwd)]
environment = {
"HOME": "/nonexistent",
"LANG": "C.UTF-8",
"LC_ALL": "C.UTF-8",
"PATH": "/usr/local/bin:/usr/bin:/bin",
"PAGER": "cat",
"RIPGREP_CONFIG_PATH": "/nonexistent",
}
try:
completed = subprocess.run(
argv,
cwd=cwd,
env=environment,
stdin=subprocess.DEVNULL,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
text=True,
errors="replace",
timeout=COMMAND_TIMEOUT_SECONDS,
check=False,
)
except subprocess.TimeoutExpired as exc:
partial = (exc.stdout or "") + (exc.stderr or "")
output, truncated = compact_output(str(partial))
return {
"program": program,
"exit_code": None,
"timed_out": True,
"truncated": truncated,
"output": output,
"instruction": "The process was killed at the fixed timeout; do not retry in a loop.",
}
output, truncated = compact_output(completed.stdout + completed.stderr)
return {
"program": program,
"exit_code": completed.returncode,
"timed_out": False,
"truncated": truncated,
"output": output,
"read_only": True,
}
def validate_source(arguments: dict[str, Any]) -> dict[str, Any]:
relative = str(arguments.get("path", ""))
target = safe_path(relative, WORKSPACE_ROOT)
if not within_root(target, WORKSPACE_ROOT) or not target.is_file():
raise PermissionError("validation is limited to files below /workspace")
kind = str(arguments.get("kind", "auto"))
suffix = target.suffix.casefold()
if kind == "auto":
if suffix == ".py":
kind = "python"
elif suffix in {".sh", ".bash"}:
kind = "shell"
elif suffix == ".json":
kind = "json"
elif suffix in {".yaml", ".yml"}:
kind = "yaml"
else:
raise ValueError("cannot infer validation kind for this file")
text = target.read_text(encoding="utf-8", errors="strict")
if len(text) > 2_000_000:
raise ValueError("source file exceeds validation limit")
if kind == "python":
ast.parse(text, filename=str(target))
elif kind == "json":
json.loads(text)
elif kind == "yaml":
import yaml
yaml.safe_load(text)
elif kind == "shell":
completed = subprocess.run(
["/bin/bash", "-n", str(target)],
env={"PATH": "/usr/bin:/bin", "LANG": "C.UTF-8"},
stdin=subprocess.DEVNULL,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
text=True,
timeout=COMMAND_TIMEOUT_SECONDS,
check=False,
)
if completed.returncode != 0:
raise ValueError(compact_output(completed.stderr)[0])
else:
raise ValueError("unsupported validation kind")
return {
"path": str(target.relative_to(WORKSPACE_ROOT)),
"kind": kind,
"valid": True,
"executed": False,
"modified": False,
}
def call_tool(name: str, arguments: dict[str, Any]) -> str:
if name == "athena_terminal_policy":
result = policy()
elif name == "athena_terminal_run":
result = run_command(arguments)
elif name == "athena_terminal_validate_source":
result = validate_source(arguments)
else:
raise ValueError(f"unknown tool: {name}")
return json_text(result)
def response(request_id: Any, result: Any = None, error: dict[str, Any] | None = None) -> None:
payload: dict[str, Any] = {"jsonrpc": "2.0", "id": request_id}
payload["error" if error is not None else "result"] = error if error is not None else result
sys.stdout.write(json_text(payload) + "\n")
sys.stdout.flush()
def handle(message: dict[str, Any]) -> None:
method = message.get("method")
request_id = message.get("id")
if method == "initialize":
response(
request_id,
{
"protocolVersion": message.get("params", {}).get("protocolVersion", "2024-11-05"),
"capabilities": {"tools": {"listChanged": False}},
"serverInfo": {"name": "mike-ai-athena-terminal", "version": SERVER_VERSION},
},
)
elif method == "tools/list":
response(request_id, {"tools": TOOLS})
elif method == "tools/call":
params = message.get("params", {})
try:
text = call_tool(str(params.get("name", "")), params.get("arguments") or {})
response(
request_id,
{
"content": [{"type": "text", "text": text}],
"structuredContent": json.loads(text),
"isError": False,
},
)
except Exception as exc:
response(
request_id,
{
"content": [{"type": "text", "text": f"ERROR: {exc}"}],
"isError": True,
},
)
elif request_id is not None:
response(request_id, error={"code": -32601, "message": f"Method not found: {method}"})
def main() -> None:
for line in sys.stdin:
try:
if line.strip():
handle(json.loads(line))
except Exception as exc:
sys.stderr.write(f"MCP input error: {exc}\n")
sys.stderr.flush()
if __name__ == "__main__":
main()
+9 -11
View File
@@ -166,22 +166,20 @@ services:
retries: 5
start_period: 10s
mcp-athena-terminal:
mcp-athena-operator:
<<: *tool-common
build:
context: .
dockerfile: Dockerfile.athena-terminal
image: mike-ai/mcp-athena-terminal:1.0.0
container_name: mike-ai-mcp-athena-terminal
dockerfile: Dockerfile.athena-operator
image: mike-ai/mcp-athena-operator:1.0.0
container_name: mike-ai-mcp-athena-operator
environment:
ATHENA_TERMINAL_WORKSPACE: /workspace
ATHENA_TERMINAL_RUNTIME: /runtime
ATHENA_OPERATOR_SOCKET: /operator/operator.sock
volumes:
# The terminal sees only versioned source and the bounded, payload-free
# runtime snapshot. It receives no Docker socket, host filesystem,
# secrets, SSH material, devices, PID namespace or egress network.
- ${PLATFORM_STACK_DIR:-/opt/mike-ai/stack}:/workspace:ro
- ${PLATFORM_CONTEXT_RUNTIME_DIR:-/var/lib/mike-ai-platform-context}:/runtime:ro
# The unprivileged MCP facade sees only the root-owned executor socket.
# Docker, source, models, Git credentials and host paths remain on the
# executor side and are reachable only through structured operations.
- /run/mike-ai-operator:/operator:ro
networks: [tools]
healthcheck:
test: ["CMD", "python", "-c", "import socket; s=socket.create_connection(('127.0.0.1',8000),2); s.close()"]
+5
View File
@@ -34,6 +34,11 @@ systemctl daemon-reload
systemctl enable --now mike-ai-platform-context-snapshot.timer
systemctl start mike-ai-platform-context-snapshot.service
# One user-facing Athena Operator MCP controls the complete local AI platform
# through a root-side structured executor. It is intentionally not a general
# shell and exposes no raw Docker socket or host paths to the MCP container.
"$MCP_DIR/../operator/install-operator.sh"
profiles=()
if [[ -s /etc/mike-ai/homeassistant-admin-mcp.env ]]; then
profiles+=(--profile homeassistant)
@@ -1,7 +1,7 @@
"""
title: MikeAI Auto Tool Selector
author: MikeAI
version: 1.1.0
version: 2.0.0
description: Selects a small, relevant set of MCP servers for each user request.
"""
@@ -27,7 +27,7 @@ class Filter:
"unraid": "server:mcp:unraid-readonly-local",
"navidrome": "server:mcp:navidrome-local",
"platform": "server:mcp:athena-platform",
"terminal": "server:mcp:athena-terminal-local",
"operator": "server:mcp:athena-operator-local",
}
LABELS = {
@@ -38,7 +38,7 @@ class Filter:
"unraid": "Unraid-Diagnose",
"navidrome": "Navidrome",
"platform": "Athena-Plattformwissen",
"terminal": "Athena-Terminal (begrenzt, nur lesend)",
"operator": "Athena Operator",
}
def __init__(self):
@@ -96,13 +96,14 @@ class Filter:
selected: list[str] = []
terminal = self._matches(
operator = self._matches(
text,
(
r"\bathena[- ]terminal\b",
r"\b(?:athena|ki[- ]host)\b.*\b(?:terminal|shell|kommando|befehl)\b",
r"\b(?:terminal|shell)\b.*\b(?:athena|ki[- ]host)\b",
r"\b(?:ls|tail|grep|rg|cat)\b.*\b(?:athena|ki[- ]host)\b",
r"\bathena[- ]operator\b",
r"\b(?:athena|ki[- ]host)\b.*\b(?:bau|[aä]nder|deploy|install|lad|test|start|stop|restart|konfigurier|aktualisier|entfern|erstell|implementier|entwickel)\w*\b",
r"\b(?:bau|[aä]nder|deploy|install|lad|test|start|stop|restart|konfigurier|aktualisier|entfern|erstell|implementier|entwickel)\w*\b.*\b(?:athena|ki[- ]host)\b",
r"\bmcp\b.*\b(?:bau\w*|implementier\w*|entwickel\w*|integrier\w*)\b",
r"\b(?:bau\w*|implementier\w*|entwickel\w*|integrier\w*)\b.*\bmcp\b",
),
)
platform = self._matches(
@@ -114,8 +115,6 @@ class Filter:
r"\b(?:disaster|bare metal)[- ]recovery\b",
r"\b(?:installations?|reinstall|setup)[- ]skript\b",
r"\bplattform(?:wissen|dokumentation)?\b",
r"\bmcp\b.*\b(?:bau\w*|implementier\w*|entwickel\w*|integrier\w*)\b",
r"\b(?:bau\w*|implementier\w*|entwickel\w*|integrier\w*)\b.*\bmcp\b",
),
)
homeassistant = self._matches(
@@ -168,8 +167,8 @@ class Filter:
# A specialist source is more precise than public web search. Platform
# wins over a generic Docker mention when Athena is explicitly named.
if terminal:
selected.append("terminal")
if operator:
selected.append("operator")
elif platform:
selected.append("platform")
elif homeassistant:
+27 -14
View File
@@ -165,6 +165,18 @@ with con:
connections = json.loads(row[0])
if not isinstance(connections, list):
raise SystemExit("Unbekanntes Format in tool_server.connections.")
before_count = len(connections)
connections = [
connection for connection in connections
if not (
isinstance(connection, dict)
and (
str((connection.get("info") or {}).get("id", "")).lower()
== "athena-terminal-local"
or "mike-ai-mcp-athena-terminal" in str(connection.get("url", "")).lower()
)
)
]
descriptions = {
"web-local": (
"Web (öffentlich, read-only)",
@@ -212,15 +224,16 @@ with con:
"Quellcode dieses Werkzeug statt allgemeiner Websuche verwenden. Keine Issues, "
"Pull Requests, Actions oder Schreibzugriffe.",
),
"athena-terminal-local": (
"Athena Terminal (begrenzt, nur lesend)",
"Nur für gezielte Beweissuche in Athenas versioniertem Stack und einem "
"begrenzten Laufzeitsnapshot. Unterstützt ausschließlich fest erlaubte "
"Lesebefehle und Syntaxprüfung. Kein SSH, Docker, Netzwerk, Schreibzugriff, "
"Reboot, Shutdown, systemctl, Interpreter oder Zugriff auf Secrets.",
"athena-operator-local": (
"Athena Operator",
"Zentrale Bedienebene für Athenas KI-Plattform: MCPs entwickeln und deployen, "
"Docker-Dienste verwalten, Modelle laden und testen, Profile/OpenWebUI ändern, "
"prüfen, dokumentieren, versionieren und Recovery erzeugen. Änderungen benötigen "
"eine inhaltlich gebundene Vorschau und ausdrückliche Bestätigung. Kein freies "
"Terminal und keine SSH-, Netzwerk-, Boot-, Reboot- oder Shutdown-Änderungen.",
),
}
changed = False
changed = len(connections) != before_count
for connection in connections:
if not isinstance(connection, dict):
continue
@@ -244,8 +257,8 @@ with con:
match = "navidrome-local"
elif "mike-ai-mcp-github" in url:
match = "github-local"
elif "mike-ai-mcp-athena-terminal" in url:
match = "athena-terminal-local"
elif "mike-ai-mcp-athena-operator" in url:
match = "athena-operator-local"
elif "mike-ai-mcp-unraid-official" in url:
match = "unraid-readonly-local"
else:
@@ -287,16 +300,16 @@ with con:
isinstance(connection, dict)
and (
str(connection.get("url", "")).lower()
== "http://mike-ai-mcp-athena-terminal:8000/mcp"
== "http://mike-ai-mcp-athena-operator:8000/mcp"
or str((connection.get("info") or {}).get("id", "")).lower()
== "athena-terminal-local"
== "athena-operator-local"
)
for connection in connections
):
name, description = descriptions["athena-terminal-local"]
name, description = descriptions["athena-operator-local"]
connections.append(
{
"url": "http://mike-ai-mcp-athena-terminal:8000/mcp",
"url": "http://mike-ai-mcp-athena-operator:8000/mcp",
"path": "",
"type": "mcp",
"auth_type": "none",
@@ -304,7 +317,7 @@ with con:
"key": "",
"config": {"enable": True, "access_grants": []},
"info": {
"id": "athena-terminal-local",
"id": "athena-operator-local",
"name": name,
"description": description,
},
+20
View File
@@ -0,0 +1,20 @@
[Unit]
Description=MikeAI Athena platform operator executor
After=docker.service data.mount network-online.target
Wants=docker.service network-online.target
[Service]
Type=simple
EnvironmentFile=-/etc/mike-ai/athena-operator.env
ExecStart=/usr/local/libexec/mike-ai-athena-operatord
Restart=on-failure
RestartSec=3
UMask=0077
NoNewPrivileges=yes
PrivateTmp=yes
ProtectHome=read-only
ProtectSystem=full
ReadWritePaths=/opt/mike-ai/stack /data/mike-ai-operator /data/models /run/mike-ai-operator /data
[Install]
WantedBy=multi-user.target
+506
View File
@@ -0,0 +1,506 @@
#!/usr/bin/env python3
"""Root-side executor for the single Athena Operator MCP.
The daemon exposes structured platform operations over a local Unix socket.
It deliberately has no arbitrary-command endpoint. Every mutation is first
materialised as an expiring, content-bound proposal and requires its exact
confirmation string in a later call.
"""
from __future__ import annotations
import difflib
import hashlib
import json
import os
import re
import shutil
import socketserver
import subprocess
import tempfile
import threading
import time
import urllib.parse
import urllib.request
import uuid
from pathlib import Path
from typing import Any
VERSION = "1.0.0"
STACK = Path(os.environ.get("ATHENA_OPERATOR_STACK", "/opt/mike-ai/stack")).resolve()
REPOSITORY = Path(os.environ.get("ATHENA_OPERATOR_REPOSITORY", "/data/mike-ai-operator/repository")).resolve()
STATE = Path(os.environ.get("ATHENA_OPERATOR_STATE", "/data/mike-ai-operator/state")).resolve()
SOCKET = Path(os.environ.get("ATHENA_OPERATOR_SOCKET", "/run/mike-ai-operator/operator.sock"))
MODELS = Path(os.environ.get("ATHENA_OPERATOR_MODELS", "/data/models")).resolve()
TICKET_TTL = 1800
MAX_FILE_BYTES = 1_000_000
MAX_FILES = 24
MAX_OUTPUT = 30_000
LOCK = threading.RLock()
SAFE_PATH = re.compile(r"^[A-Za-z0-9_.+/-]{1,240}$")
SAFE_NAME = re.compile(r"^[A-Za-z0-9_.-]{1,100}$")
SAFE_COMMIT = re.compile(r"^[A-Za-z0-9ÄÖÜäöüß _.,:;()+/\-]{5,120}$")
BLOCKED_PATH_PARTS = {".git", ".ssh", "secrets", "credentials", "authorized_keys"}
PROTECTED_CONTAINERS = {
"mike-ai-wireguard-gateway",
}
ALLOWED_OPERATIONS = {
"file_update", "run_checks", "compose_deploy", "container_action",
"openwebui_sync", "git_publish", "model_download", "benchmark", "recovery",
}
ALLOWED_CHECKS = {
"operator-tests": ["python3", "dev/test_athena_operator.py"],
"openwebui-filter-tests": ["python3", "dev/test_openwebui_filters.py"],
"platform-verify": ["bash", "platform/checks/verify-platform.sh"],
"compose-main": ["docker", "compose", "-f", "compose.yaml", "config", "-q"],
"compose-mcp": ["docker", "compose", "-f", "platform/mcp/compose.yaml", "config", "-q"],
}
def now() -> int:
return int(time.time())
def compact(text: str) -> str:
if len(text) <= MAX_OUTPUT:
return text
return text[:22_000] + f"\n...[truncated from {len(text)} chars]...\n" + text[-7_000:]
def run(argv: list[str], *, cwd: Path = STACK, timeout: int = 900, check: bool = False) -> dict[str, Any]:
completed = subprocess.run(
argv, cwd=cwd, env={"PATH": "/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin", "LANG": "C.UTF-8", "HOME": "/root"},
stdin=subprocess.DEVNULL, stdout=subprocess.PIPE, stderr=subprocess.STDOUT,
text=True, errors="replace", timeout=timeout, check=False,
)
result = {"argv": argv, "exit_code": completed.returncode, "output": compact(completed.stdout)}
if check and completed.returncode != 0:
raise RuntimeError(json.dumps(result, ensure_ascii=False))
return result
def safe_relative(value: str) -> Path:
if not SAFE_PATH.fullmatch(value or "") or value.startswith("/"):
raise ValueError("invalid repository-relative path")
path = Path(value)
if ".." in path.parts or any(part.casefold() in BLOCKED_PATH_PARTS for part in path.parts):
raise PermissionError("protected path")
resolved = (REPOSITORY / path).resolve(strict=False)
resolved.relative_to(REPOSITORY)
return path
def source_file(root: Path, relative: Path) -> Path:
candidate = (root / relative).resolve(strict=False)
candidate.relative_to(root)
return candidate
def sha(data: bytes) -> str:
return hashlib.sha256(data).hexdigest()
def json_write(path: Path, value: Any) -> None:
path.parent.mkdir(parents=True, exist_ok=True)
temporary = path.with_suffix(path.suffix + ".tmp")
temporary.write_text(json.dumps(value, ensure_ascii=False, indent=2), encoding="utf-8")
os.replace(temporary, path)
def audit(event: str, **fields: Any) -> None:
record = {"time": now(), "event": event, **fields}
path = STATE / "audit.jsonl"
path.parent.mkdir(parents=True, exist_ok=True)
with path.open("a", encoding="utf-8") as handle:
handle.write(json.dumps(record, ensure_ascii=False, separators=(",", ":")) + "\n")
def ensure_repository() -> None:
if not (REPOSITORY / ".git").is_dir():
bundle = Path("/data/mike-ai-recovery-kit/source.git.bundle")
if not bundle.is_file():
raise RuntimeError("operator repository missing and no recovery Git bundle is available")
REPOSITORY.parent.mkdir(parents=True, exist_ok=True)
run(["git", "clone", str(bundle), str(REPOSITORY)], cwd=Path("/data"), check=True)
current_file = STACK / ".mike-ai-source-commit"
current = current_file.read_text().strip() if current_file.is_file() else ""
if re.fullmatch(r"[0-9a-f]{40}", current):
run(["git", "switch", "-C", "main", current], cwd=REPOSITORY, check=True)
remote = os.environ.get("ATHENA_OPERATOR_GIT_REMOTE", "").strip()
if remote:
run(["git", "remote", "set-url", "origin", remote], cwd=REPOSITORY, check=True)
run(["git", "config", "user.name", os.environ.get("ATHENA_OPERATOR_GIT_NAME", "Athena Operator")], cwd=REPOSITORY, check=True)
run(["git", "config", "user.email", os.environ.get("ATHENA_OPERATOR_GIT_EMAIL", "athena-operator@localhost")], cwd=REPOSITORY, check=True)
def inspect(subject: str, arguments: dict[str, Any]) -> dict[str, Any]:
ensure_repository()
if subject == "overview":
return {
"version": VERSION,
"source_commit": (STACK / ".mike-ai-source-commit").read_text().strip(),
"git": run(["git", "status", "--short", "--branch"], cwd=REPOSITORY),
"containers": run(["docker", "ps", "--format", "{{.Names}}\t{{.Status}}\t{{.Image}}"]),
"storage": run(["df", "-h", "/", "/data", str(MODELS)]),
"gpus": run(["nvidia-smi", "--query-gpu=name,memory.total,memory.used,utilization.gpu", "--format=csv,noheader"]),
"boundary": "Athena AI-platform operator; no arbitrary shell, shutdown, reboot, SSH/network/firewall/kernel/driver/partition operations",
}
if subject == "git_status":
return {"status": run(["git", "status", "--short", "--branch"], cwd=REPOSITORY), "diff": run(["git", "diff", "--stat"], cwd=REPOSITORY)}
if subject == "containers":
return {"containers": run(["docker", "ps", "-a", "--format", "{{.Names}}\t{{.Status}}\t{{.Image}}"])}
if subject == "models":
entries = []
if MODELS.is_dir():
for path in sorted(MODELS.rglob("*.gguf")):
entries.append({"path": str(path.relative_to(MODELS)), "bytes": path.stat().st_size})
return {"models": entries[:500], "count": len(entries)}
if subject == "jobs":
jobs = []
for path in sorted((STATE / "jobs").glob("*.json"), reverse=True)[:30]:
jobs.append(json.loads(path.read_text()))
return {"jobs": jobs}
raise ValueError("unsupported inspection subject")
def read_source(arguments: dict[str, Any]) -> dict[str, Any]:
ensure_repository()
relative = safe_relative(str(arguments.get("path", "")))
target = source_file(REPOSITORY, relative)
if not target.is_file():
raise FileNotFoundError("source file not found")
text = target.read_text(encoding="utf-8", errors="replace")
start = max(1, int(arguments.get("start_line", 1)))
count = min(1000, max(1, int(arguments.get("line_count", 300))))
lines = text.splitlines()
return {"path": str(relative), "sha256": sha(target.read_bytes()), "start_line": start, "content": "\n".join(lines[start - 1:start - 1 + count]), "total_lines": len(lines)}
def search_source(arguments: dict[str, Any]) -> dict[str, Any]:
ensure_repository()
query = str(arguments.get("query", ""))
if not query or len(query) > 200 or any(x in query for x in ("\x00", "\n", "\r")):
raise ValueError("invalid query")
result = run(["rg", "-n", "--hidden", "--glob", "!.git/**", "--", query, "."], cwd=REPOSITORY, timeout=20)
return {"query": query, "matches": result["output"], "exit_code": result["exit_code"]}
def normalise_operation(operation: str, payload: dict[str, Any]) -> tuple[dict[str, Any], str]:
if operation not in ALLOWED_OPERATIONS:
raise ValueError("unsupported operation")
if not isinstance(payload, dict):
raise ValueError("payload must be an object")
if operation == "file_update":
files = payload.get("files")
if not isinstance(files, list) or not 1 <= len(files) <= MAX_FILES:
raise ValueError("files must contain 1..24 entries")
normal = []
previews = []
for item in files:
relative = safe_relative(str(item.get("path", "")))
content = str(item.get("content", ""))
raw = content.encode()
if len(raw) > MAX_FILE_BYTES:
raise ValueError("file content exceeds limit")
target = source_file(REPOSITORY, relative)
before = target.read_text(encoding="utf-8", errors="replace") if target.is_file() else ""
expected = str(item.get("expected_sha256", ""))
before_sha = sha(before.encode())
if expected and expected != before_sha:
raise RuntimeError(f"source drift for {relative}")
diff = "".join(difflib.unified_diff(before.splitlines(True), content.splitlines(True), fromfile=f"a/{relative}", tofile=f"b/{relative}"))
normal.append({"path": str(relative), "content": content, "before_sha256": before_sha, "after_sha256": sha(raw)})
previews.append(compact(diff))
return {"files": normal}, "\n".join(previews)
if operation == "run_checks":
checks = payload.get("checks") or []
if not isinstance(checks, list) or not checks or any(name not in ALLOWED_CHECKS for name in checks):
raise ValueError("unknown check suite")
return {"checks": checks}, "Will run: " + ", ".join(checks)
if operation == "compose_deploy":
compose_file = str(payload.get("compose_file", ""))
if compose_file not in {"compose.yaml", "platform/mcp/compose.yaml"}:
raise ValueError("unsupported compose file")
services = payload.get("services") or []
if not isinstance(services, list) or not 1 <= len(services) <= 12 or any(not SAFE_NAME.fullmatch(str(x)) for x in services):
raise ValueError("invalid compose service list")
return {"compose_file": compose_file, "services": services, "build": bool(payload.get("build", True))}, f"docker compose -f {compose_file} up -d {'--build ' if payload.get('build', True) else ''}{' '.join(services)}"
if operation == "container_action":
action = str(payload.get("action", ""))
containers = payload.get("containers") or []
if action not in {"start", "stop", "restart"}:
raise ValueError("invalid container action")
if not isinstance(containers, list) or not 1 <= len(containers) <= 12:
raise ValueError("invalid container list")
for name in containers:
if not SAFE_NAME.fullmatch(str(name)) or not str(name).startswith("mike-ai-") or name in PROTECTED_CONTAINERS:
raise PermissionError(f"container is outside operator boundary: {name}")
return {"action": action, "containers": containers}, f"docker {action} {' '.join(containers)}"
if operation == "openwebui_sync":
return {}, (
"Synchronise the versioned Open WebUI model profiles, filters, tool connections "
"and system prompt with the running Open WebUI instance."
)
if operation == "git_publish":
message = str(payload.get("message", ""))
if not SAFE_COMMIT.fullmatch(message):
raise ValueError("invalid commit message")
status = run(["git", "status", "--short"], cwd=REPOSITORY, check=True)["output"].strip()
if not status:
raise RuntimeError("repository has no changes to publish")
diff = run(["git", "diff", "--stat"], cwd=REPOSITORY, check=True)["output"]
preview = f"Commit message: {message}\n\nChanged and untracked files:\n{status}\n\nDiff summary:\n{diff}"
return {"message": message, "reviewed_status": status}, preview
if operation == "model_download":
url = str(payload.get("url", ""))
parsed = urllib.parse.urlparse(url)
if parsed.scheme != "https" or parsed.hostname not in {"huggingface.co", "cdn-lfs.huggingface.co", "hf.co"}:
raise PermissionError("only HTTPS Hugging Face downloads are allowed")
destination = safe_relative(str(payload.get("destination", "")))
expected = str(payload.get("sha256", ""))
if expected and not re.fullmatch(r"[0-9a-f]{64}", expected):
raise ValueError("invalid sha256")
return {"url": url, "destination": str(destination), "sha256": expected}, f"Download {url} to models/{destination}"
if operation == "benchmark":
script = safe_relative(str(payload.get("script", "")))
if not (str(script).startswith("dev/") or str(script).startswith("platform/bench")) or script.suffix not in {".py", ".sh"}:
raise PermissionError("benchmark script must be versioned below dev/ or platform/bench*")
args = payload.get("arguments") or []
if not isinstance(args, list) or len(args) > 20 or any(not isinstance(x, str) or len(x) > 200 or re.search(r"[\x00\n\r]", x) for x in args):
raise ValueError("invalid benchmark arguments")
return {"script": str(script), "arguments": args}, f"Run versioned benchmark {script} with {args!r}"
if operation == "recovery":
label = str(payload.get("label", time.strftime("%Y%m%d")))
if not SAFE_NAME.fullmatch(label):
raise ValueError("invalid recovery label")
return {"label": label}, f"Create encrypted recovery bundle and self-contained data kit: {label}"
raise AssertionError(operation)
def prepare(arguments: dict[str, Any]) -> dict[str, Any]:
ensure_repository()
operation = str(arguments.get("operation", ""))
payload, preview = normalise_operation(operation, arguments.get("payload") or {})
ticket_id = uuid.uuid4().hex
record = {"id": ticket_id, "operation": operation, "payload": payload, "preview": preview, "created": now(), "expires": now() + TICKET_TTL, "status": "pending"}
record["binding"] = sha(json.dumps({"operation": operation, "payload": payload}, sort_keys=True, ensure_ascii=False).encode())
json_write(STATE / "pending" / f"{ticket_id}.json", record)
audit("prepared", ticket=ticket_id, operation=operation, binding=record["binding"])
return {"ticket": ticket_id, "operation": operation, "binding": record["binding"], "preview": preview, "expires_in_seconds": TICKET_TTL, "required_confirmation": f"EXECUTE {ticket_id}", "instruction": "Show the full preview to the user and wait for explicit confirmation in a later message."}
def sync_file(relative: Path, content: str, backup_root: Path) -> None:
for root in (REPOSITORY, STACK):
target = source_file(root, relative)
if target.exists():
backup = backup_root / root.name / relative
backup.parent.mkdir(parents=True, exist_ok=True)
shutil.copy2(target, backup)
target.parent.mkdir(parents=True, exist_ok=True)
fd, temp_name = tempfile.mkstemp(prefix=f".{target.name}.", dir=target.parent)
with os.fdopen(fd, "w", encoding="utf-8") as handle:
handle.write(content)
os.replace(temp_name, target)
def start_job(ticket: str, operation: str, worker) -> dict[str, Any]:
job_id = uuid.uuid4().hex
job_path = STATE / "jobs" / f"{job_id}.json"
json_write(job_path, {"id": job_id, "ticket": ticket, "operation": operation, "status": "running", "started": now()})
def wrapped():
try:
result = worker()
json_write(job_path, {"id": job_id, "ticket": ticket, "operation": operation, "status": "completed", "started": json.loads(job_path.read_text())["started"], "finished": now(), "result": result})
audit("job_completed", job=job_id, operation=operation)
except Exception as exc:
json_write(job_path, {"id": job_id, "ticket": ticket, "operation": operation, "status": "failed", "finished": now(), "error": compact(str(exc))})
audit("job_failed", job=job_id, operation=operation)
threading.Thread(target=wrapped, daemon=True).start()
return {"job_id": job_id, "status": "running", "instruction": "Poll athena_operator_job until completed or failed."}
def execute_operation(ticket: str, operation: str, payload: dict[str, Any]) -> dict[str, Any]:
if operation == "file_update":
backup = STATE / "backups" / f"{now()}-{ticket}"
for item in payload["files"]:
relative = safe_relative(item["path"])
current = source_file(REPOSITORY, relative)
current_sha = sha(current.read_bytes()) if current.is_file() else sha(b"")
if current_sha != item["before_sha256"]:
raise RuntimeError(f"source drift after preview: {relative}")
for item in payload["files"]:
sync_file(safe_relative(item["path"]), item["content"], backup)
return {"changed": [item["path"] for item in payload["files"]], "backup": str(backup), "git_diff": run(["git", "diff", "--stat"], cwd=REPOSITORY)}
if operation == "run_checks":
return {"checks": [{"name": name, **run(ALLOWED_CHECKS[name], cwd=REPOSITORY, timeout=1200)} for name in payload["checks"]]}
if operation == "compose_deploy":
compose = payload["compose_file"]
run(["docker", "compose", "-f", compose, "config", "-q"], cwd=STACK, check=True)
argv = ["docker", "compose", "-f", compose, "up", "-d"]
if payload["build"]:
argv.append("--build")
argv.extend(payload["services"])
return {"deploy": run(argv, cwd=STACK, timeout=3600, check=True), "containers": run(["docker", "ps", "--format", "{{.Names}}\t{{.Status}}"])}
if operation == "container_action":
return {"action": run(["docker", payload["action"], *payload["containers"]], timeout=300, check=True)}
if operation == "openwebui_sync":
installer = STACK / "platform/openwebui/install-filters.sh"
if not installer.is_file():
raise FileNotFoundError("versioned Open WebUI synchronisation script is missing")
return {"sync": run(["bash", str(installer)], cwd=STACK, timeout=1800, check=True)}
if operation == "git_publish":
current_status = run(["git", "status", "--short"], cwd=REPOSITORY, check=True)["output"].strip()
if current_status != payload["reviewed_status"]:
raise RuntimeError("repository changed after the Git publish preview")
run(["git", "add", "--all"], cwd=REPOSITORY, check=True)
run(["git", "diff", "--cached", "--check"], cwd=REPOSITORY, check=True)
commit = run(["git", "commit", "-m", payload["message"]], cwd=REPOSITORY, check=True)
pushed = run(["git", "push", "origin", "HEAD:main"], cwd=REPOSITORY, timeout=300, check=True)
head = run(["git", "rev-parse", "HEAD"], cwd=REPOSITORY, check=True)["output"].strip()
(STACK / ".mike-ai-source-commit").write_text(head + "\n")
return {"commit": head, "commit_output": commit, "push_output": pushed}
if operation == "model_download":
def download():
destination = source_file(MODELS, Path(payload["destination"]))
destination.parent.mkdir(parents=True, exist_ok=True)
partial = destination.with_suffix(destination.suffix + ".partial")
request = urllib.request.Request(payload["url"], headers={"User-Agent": "Athena-Operator/1"})
digest = hashlib.sha256()
total = 0
with urllib.request.urlopen(request, timeout=60) as response, partial.open("wb") as output:
while chunk := response.read(8 * 1024 * 1024):
output.write(chunk); digest.update(chunk); total += len(chunk)
actual = digest.hexdigest()
if payload["sha256"] and actual != payload["sha256"]:
partial.unlink(missing_ok=True); raise RuntimeError("model checksum mismatch")
os.replace(partial, destination)
return {"destination": str(destination), "bytes": total, "sha256": actual}
return start_job(ticket, operation, download)
if operation == "benchmark":
def benchmark():
script = source_file(REPOSITORY, safe_relative(payload["script"]))
interpreter = "python3" if script.suffix == ".py" else "bash"
return run([interpreter, str(script), *payload["arguments"]], cwd=REPOSITORY, timeout=86400)
return start_job(ticket, operation, benchmark)
if operation == "recovery":
def recovery():
label = payload["label"]
dirty = run(["git", "status", "--porcelain"], cwd=REPOSITORY, check=True)["output"].strip()
if dirty:
raise RuntimeError("publish repository changes before creating a recovery kit")
head = run(["git", "rev-parse", "HEAD"], cwd=REPOSITORY, check=True)["output"].strip()
marker = (STACK / ".mike-ai-source-commit").read_text().strip()
if marker != head:
raise RuntimeError("deployed source marker and operator repository HEAD differ")
encrypted = Path(f"/data/athena-recovery-{label}.tar.age")
source_bundle = Path(f"/data/athena-source-{label}.git.bundle")
release = Path(f"/data/mike-ai-recovery-kit-{label}")
for target in (encrypted, source_bundle, release):
if target.exists():
raise FileExistsError(target)
git_bundle = run(["git", "bundle", "create", str(source_bundle), "--all"], cwd=REPOSITORY, timeout=1800, check=True)
encrypted_result = run([str(STACK / "platform/recovery/create-recovery-bundle.sh"), str(encrypted)], timeout=7200, check=True)
identity = Path("/data/mike-ai-recovery-kit/recovery.agekey")
if not identity.is_file():
raise RuntimeError("existing recovery identity is unavailable")
kit_result = run([
str(STACK / "platform/recovery/create-self-contained-data-kit.sh"),
str(encrypted), str(identity), str(source_bundle), str(release),
], timeout=7200, check=True)
verify = run(["sha256sum", "-c", "SHA256SUMS"], cwd=release, timeout=1800, check=True)
return {
"commit": head,
"recovery_bundle": str(encrypted),
"source_bundle": str(source_bundle),
"self_contained_kit": str(release),
"git_bundle": git_bundle,
"encrypted_bundle": encrypted_result,
"kit": kit_result,
"verification": verify,
}
return start_job(ticket, operation, recovery)
raise AssertionError(operation)
def execute(arguments: dict[str, Any]) -> dict[str, Any]:
ticket = str(arguments.get("ticket", ""))
confirmation = str(arguments.get("confirmation", ""))
if not re.fullmatch(r"[0-9a-f]{32}", ticket) or confirmation != f"EXECUTE {ticket}":
raise PermissionError("exact ticket confirmation required")
path = STATE / "pending" / f"{ticket}.json"
if not path.is_file():
raise ValueError("ticket not found or already consumed")
with LOCK:
record = json.loads(path.read_text())
if record["status"] != "pending" or now() > record["expires"]:
raise PermissionError("ticket expired or unavailable")
expected = sha(json.dumps({"operation": record["operation"], "payload": record["payload"]}, sort_keys=True, ensure_ascii=False).encode())
if expected != record["binding"]:
raise RuntimeError("ticket binding mismatch")
record["status"] = "executing"
json_write(path, record)
try:
result = execute_operation(ticket, record["operation"], record["payload"])
record["status"] = "executed"
record["executed"] = now()
record["result_summary"] = compact(json.dumps(result, ensure_ascii=False))
completed = STATE / "completed" / path.name
json_write(completed, record)
path.unlink()
audit("executed", ticket=ticket, operation=record["operation"], binding=record["binding"])
return {"ticket": ticket, "operation": record["operation"], "result": result, "instruction": "Verify health and Git/recovery state before declaring the work complete."}
except Exception:
record["status"] = "failed"
json_write(path, record)
raise
def job(arguments: dict[str, Any]) -> dict[str, Any]:
job_id = str(arguments.get("job_id", ""))
if not re.fullmatch(r"[0-9a-f]{32}", job_id):
raise ValueError("invalid job id")
path = STATE / "jobs" / f"{job_id}.json"
if not path.is_file():
raise FileNotFoundError("job not found")
return json.loads(path.read_text())
def dispatch(request: dict[str, Any]) -> dict[str, Any]:
action = request.get("action")
arguments = request.get("arguments") or {}
if action == "inspect": return inspect(str(arguments.get("subject", "overview")), arguments)
if action == "read_source": return read_source(arguments)
if action == "search_source": return search_source(arguments)
if action == "prepare": return prepare(arguments)
if action == "execute": return execute(arguments)
if action == "job": return job(arguments)
raise ValueError("unsupported operator action")
class Handler(socketserver.StreamRequestHandler):
def handle(self) -> None:
line = self.rfile.readline(2_000_000)
try:
request = json.loads(line)
result = dispatch(request)
response = {"ok": True, "result": result}
except Exception as exc:
response = {"ok": False, "error": compact(str(exc))}
self.wfile.write((json.dumps(response, ensure_ascii=False, separators=(",", ":")) + "\n").encode())
def main() -> None:
STATE.mkdir(parents=True, exist_ok=True)
SOCKET.parent.mkdir(parents=True, exist_ok=True)
SOCKET.unlink(missing_ok=True)
server = socketserver.ThreadingUnixStreamServer(str(SOCKET), Handler)
os.chmod(SOCKET, 0o660)
os.chown(SOCKET, 0, int(os.environ.get("ATHENA_OPERATOR_GID", "10003")))
audit("daemon_started", version=VERSION)
server.serve_forever()
if __name__ == "__main__":
main()
+16
View File
@@ -0,0 +1,16 @@
#!/usr/bin/env bash
set -Eeuo pipefail
[[ $EUID -eq 0 ]] || { echo "Bitte als root ausführen." >&2; exit 1; }
ROOT=$(cd "$(dirname "${BASH_SOURCE[0]}")/../.." && pwd)
install -d -m 0700 /etc/mike-ai
if [[ ! -e /etc/mike-ai/athena-operator.env ]]; then
install -m 0600 "$ROOT/config/athena-operator.env.example" /etc/mike-ai/athena-operator.env
fi
install -d -m 0750 -o root -g 10003 /run/mike-ai-operator
install -d -m 0700 /data/mike-ai-operator/state /data/mike-ai-operator/repository
install -m 0755 "$ROOT/platform/operator/athena_operatord.py" /usr/local/libexec/mike-ai-athena-operatord
install -m 0644 "$ROOT/platform/operator/athena-operator.service" /etc/systemd/system/mike-ai-athena-operator.service
systemctl daemon-reload
systemctl enable --now mike-ai-athena-operator.service
systemctl is-active --quiet mike-ai-athena-operator.service
echo ATHENA_OPERATOR_EXECUTOR_OK