mirror of
https://github.com/pewdiepie-archdaemon/odysseus.git
synced 2026-06-16 17:55:26 -04:00
feat: add ChatGPT Subscription provider (#2876)
* feat: Add ChatGPT Subscription support and related features - Introduced a new provider option for ChatGPT Subscription in the endpoint selection UI. - Implemented OAuth flow for ChatGPT Subscription sign-in, including polling for authorization status. - Updated admin interface to handle ChatGPT Subscription, including disabling API key input and providing user guidance. - Enhanced cost tracking logic to differentiate between subscription and non-subscription endpoints. - Added new slash commands for managing skills, including listing, searching, and invoking skills. - Implemented caching for skill catalog to optimize performance. - Updated tests to cover new ChatGPT Subscription functionality and ensure proper endpoint probing. - Refactored existing code to accommodate new features and improve maintainability. * refactor: share provider device-flow setup - reuse one device-flow backend for Copilot and ChatGPT Subscription - add one frontend device-flow helper for Settings and /setup - put GitHub Copilot back into Add Models, now as a dropdown option - make provider selection just select; clicking Add starts sign-in - stop ChatGPT Subscription setup from opening auth tabs automatically - make /setup copilot and /setup chatgpt-subscription work from chat - show ChatGPT Subscription in the /setup suggestions - show the real error message when setup fails - add focused tests for the shared flow and setup UI * feat(chatgpt-subscription): harden credential lifecycle and streamline auth UX Backend: - Resolve runtime bearer for provider-auth endpoints at probe time via a shared _resolve_probe_key() that delegates to resolve_endpoint_runtime, applied across all probe/refresh call sites. - Skip live completion probes and health pings for discovery-only providers (centralized behind _is_discovery_only_provider) — the Codex/Responses API has no such endpoints, so status is derived from cached models. - Never persist the short lived ChatGPT bearer to the plaintext sessions table; proactively clear any stale bearer left by an earlier code path. - Revoke orphaned ProviderAuthSession credentials when the last endpoint backing them is deleted (_delete_orphaned_provider_auth), surfaced via cleared_provider_auth in the delete response. Frontend (admin.js): - Auto-start the device-auth flow on provider selection so the authorization panel (code + Authorize) shows immediately instead of behind a "Sign in" click. - Remove the redundant top button for device auth providers, move retry into the panel via an inline "Try again". - Drop the self-evident hint text and add an execCommand clipboard fallback so Copy works in non-secure (HTTP/LAN) contexts. * fix: harden chatgpt subscription provider * chore: remove PR media from branch * Fix chatgpt subscription recovery and token handling --------- Co-authored-by: 5p00kyy <admin@5p00ky.dev>
This commit is contained in:
stocky789
+217
-7
@@ -426,6 +426,9 @@ def _detect_provider(url: str) -> str:
|
||||
return "openrouter"
|
||||
if _host_match(url, "groq.com"):
|
||||
return "groq"
|
||||
from src.chatgpt_subscription import is_chatgpt_subscription_base
|
||||
if is_chatgpt_subscription_base(url):
|
||||
return "chatgpt-subscription"
|
||||
from src.copilot import is_copilot_base
|
||||
if is_copilot_base(url):
|
||||
return "copilot"
|
||||
@@ -462,6 +465,8 @@ def _provider_label(url: str) -> str:
|
||||
if _host_match(url, "opencode.ai/zen/go"): return "OpenCode Go"
|
||||
if _host_match(url, "opencode.ai/zen"): return "OpenCode Zen"
|
||||
if _host_match(url, "groq.com"): return "Groq"
|
||||
from src.chatgpt_subscription import is_chatgpt_subscription_base
|
||||
if is_chatgpt_subscription_base(url): return "ChatGPT Subscription"
|
||||
from src.copilot import is_copilot_base
|
||||
if is_copilot_base(url): return "GitHub Copilot"
|
||||
if _host_match(url, "mistral.ai"): return "Mistral"
|
||||
@@ -479,6 +484,77 @@ def _provider_label(url: str) -> str:
|
||||
return host or "provider"
|
||||
|
||||
|
||||
def _normalize_chatgpt_subscription_url(url: str) -> str:
|
||||
base = (url or "").strip().rstrip("/")
|
||||
if base.endswith("/responses"):
|
||||
return base
|
||||
return base + "/responses"
|
||||
|
||||
|
||||
def _message_content_as_text(content) -> str:
|
||||
if isinstance(content, str):
|
||||
return content
|
||||
if isinstance(content, list):
|
||||
parts: list[str] = []
|
||||
for part in content:
|
||||
if not isinstance(part, dict):
|
||||
if part:
|
||||
parts.append(str(part))
|
||||
continue
|
||||
if isinstance(part.get("text"), str):
|
||||
parts.append(part["text"])
|
||||
continue
|
||||
if isinstance(part.get("content"), str):
|
||||
parts.append(part["content"])
|
||||
return "\n".join(parts)
|
||||
return "" if content is None else str(content)
|
||||
|
||||
|
||||
def _chatgpt_subscription_instructions(messages: List[Dict]) -> str:
|
||||
instructions = [
|
||||
_message_content_as_text(msg.get("content")).strip()
|
||||
for msg in messages or []
|
||||
if (msg.get("role") or "") == "system"
|
||||
]
|
||||
instructions = [part for part in instructions if part]
|
||||
if instructions:
|
||||
return "\n\n".join(instructions)
|
||||
return "You are a helpful AI assistant."
|
||||
|
||||
|
||||
def _build_chatgpt_responses_payload(
|
||||
model: str,
|
||||
messages: List[Dict],
|
||||
temperature: float,
|
||||
max_tokens: int,
|
||||
*,
|
||||
stream: bool = False,
|
||||
) -> Dict:
|
||||
from src.chatgpt_subscription import build_responses_input
|
||||
|
||||
conversation = [msg for msg in (messages or []) if (msg.get("role") or "") != "system"]
|
||||
payload: Dict = {
|
||||
"model": model,
|
||||
"instructions": _chatgpt_subscription_instructions(messages),
|
||||
"input": build_responses_input(conversation),
|
||||
"stream": stream,
|
||||
"store": False,
|
||||
}
|
||||
if not _restricts_temperature(model):
|
||||
payload["temperature"] = temperature
|
||||
if max_tokens and max_tokens > 0:
|
||||
payload["max_output_tokens"] = max_tokens
|
||||
return payload
|
||||
|
||||
|
||||
def _format_chatgpt_subscription_error(status_code: int, text: str) -> str:
|
||||
if status_code in (401, 403):
|
||||
return "ChatGPT Subscription credentials expired or were rejected. Reconnect the provider."
|
||||
if status_code == 429:
|
||||
return "ChatGPT Subscription quota or rate limit was reached. Retry after the upstream limit resets."
|
||||
return _format_upstream_error(status_code, text, "https://chatgpt.com/backend-api/codex")
|
||||
|
||||
|
||||
def _format_upstream_error(status: int, body: bytes | str, url: str) -> str:
|
||||
"""Turn an upstream HTTP error into a user-readable sentence.
|
||||
|
||||
@@ -874,7 +950,7 @@ def _normalize_anthropic_url(url: str) -> str:
|
||||
def _model_list_base(url: str) -> str:
|
||||
"""Normalize model/chat URLs to the configured endpoint base."""
|
||||
base = (url or "").strip().rstrip("/")
|
||||
for suffix in ("/models", "/chat/completions", "/completions", "/v1/messages"):
|
||||
for suffix in ("/models", "/chat/completions", "/completions", "/v1/messages", "/responses"):
|
||||
if base.endswith(suffix):
|
||||
base = base[: -len(suffix)].rstrip("/")
|
||||
for suffix in ("/chat", "/tags", "/generate"):
|
||||
@@ -903,7 +979,12 @@ def _parse_model_cache(raw) -> List[str]:
|
||||
return out
|
||||
|
||||
|
||||
def _configured_cached_model_ids(endpoint_url: str) -> List[str]:
|
||||
def _configured_cached_model_ids(
|
||||
endpoint_url: str,
|
||||
*,
|
||||
owner: Optional[str] = None,
|
||||
endpoint_id: Optional[str] = None,
|
||||
) -> List[str]:
|
||||
"""Return cached models for a configured endpoint matching endpoint_url."""
|
||||
target = _model_list_base(endpoint_url)
|
||||
if not target:
|
||||
@@ -914,7 +995,13 @@ def _configured_cached_model_ids(endpoint_url: str) -> List[str]:
|
||||
return []
|
||||
db = SessionLocal()
|
||||
try:
|
||||
rows = db.query(ModelEndpoint).filter(ModelEndpoint.is_enabled == True).all()
|
||||
q = db.query(ModelEndpoint).filter(ModelEndpoint.is_enabled == True)
|
||||
if endpoint_id:
|
||||
q = q.filter(ModelEndpoint.id == endpoint_id)
|
||||
if owner:
|
||||
from src.auth_helpers import owner_filter
|
||||
q = owner_filter(q, ModelEndpoint, owner)
|
||||
rows = q.all()
|
||||
for ep in rows:
|
||||
if _model_list_base(getattr(ep, "base_url", "")) != target:
|
||||
continue
|
||||
@@ -933,9 +1020,16 @@ def _configured_cached_model_ids(endpoint_url: str) -> List[str]:
|
||||
return []
|
||||
|
||||
|
||||
def list_model_ids(base_chat_url: str, timeout: int = LLMConfig.DEFAULT_TIMEOUT, headers: Optional[Dict] = None) -> List[str]:
|
||||
def list_model_ids(
|
||||
base_chat_url: str,
|
||||
timeout: int = LLMConfig.DEFAULT_TIMEOUT,
|
||||
headers: Optional[Dict] = None,
|
||||
*,
|
||||
owner: Optional[str] = None,
|
||||
endpoint_id: Optional[str] = None,
|
||||
) -> List[str]:
|
||||
"""List available model IDs from an endpoint."""
|
||||
cached = _configured_cached_model_ids(base_chat_url)
|
||||
cached = _configured_cached_model_ids(base_chat_url, owner=owner, endpoint_id=endpoint_id)
|
||||
if cached:
|
||||
return cached
|
||||
provider = _detect_provider(base_chat_url)
|
||||
@@ -971,9 +1065,16 @@ def list_model_ids(base_chat_url: str, timeout: int = LLMConfig.DEFAULT_TIMEOUT,
|
||||
pass
|
||||
return []
|
||||
|
||||
def normalize_model_id(endpoint_url: str, requested: str, timeout: int = LLMConfig.DEFAULT_TIMEOUT) -> Optional[str]:
|
||||
def normalize_model_id(
|
||||
endpoint_url: str,
|
||||
requested: str,
|
||||
timeout: int = LLMConfig.DEFAULT_TIMEOUT,
|
||||
*,
|
||||
owner: Optional[str] = None,
|
||||
endpoint_id: Optional[str] = None,
|
||||
) -> Optional[str]:
|
||||
"""Normalize a model ID to match available models."""
|
||||
avail = list_model_ids(endpoint_url, timeout)
|
||||
avail = list_model_ids(endpoint_url, timeout, owner=owner, endpoint_id=endpoint_id)
|
||||
if not avail:
|
||||
return None
|
||||
if requested in avail:
|
||||
@@ -1169,6 +1270,49 @@ async def llm_call_async(
|
||||
logger.debug(f"Returning cached response for key: {cache_key}")
|
||||
return cached_response
|
||||
|
||||
if provider == "chatgpt-subscription":
|
||||
# ChatGPT/Codex requires streamed Responses requests even for callers
|
||||
# that want a plain string (auto-title, memory extraction, etc.).
|
||||
# Reuse stream_llm's validated Codex SSE path and collect deltas.
|
||||
parts: List[str] = []
|
||||
async for chunk in stream_llm(
|
||||
url,
|
||||
model,
|
||||
messages_copy,
|
||||
temperature=temperature,
|
||||
max_tokens=max_tokens,
|
||||
headers=headers,
|
||||
timeout=timeout,
|
||||
):
|
||||
event_is_error = False
|
||||
for line in str(chunk).splitlines():
|
||||
if line.startswith("event:"):
|
||||
event_is_error = line[6:].strip() == "error"
|
||||
continue
|
||||
if not line.startswith("data:"):
|
||||
continue
|
||||
raw = line[5:].strip()
|
||||
if not raw:
|
||||
continue
|
||||
if raw == "[DONE]":
|
||||
response = "".join(parts)
|
||||
_set_cached_response(cache_key, response)
|
||||
return response
|
||||
try:
|
||||
data = json.loads(raw)
|
||||
except json.JSONDecodeError:
|
||||
continue
|
||||
if event_is_error or data.get("error") or (data.get("status") and data.get("text")):
|
||||
status = int(data.get("status") or 502)
|
||||
text = data.get("text") or data.get("error") or "ChatGPT Subscription request failed"
|
||||
raise HTTPException(status, text)
|
||||
delta = data.get("delta")
|
||||
if isinstance(delta, str):
|
||||
parts.append(delta)
|
||||
response = "".join(parts)
|
||||
_set_cached_response(cache_key, response)
|
||||
return response
|
||||
|
||||
if provider == "anthropic":
|
||||
target_url = _normalize_anthropic_url(url)
|
||||
h = _build_anthropic_headers(headers)
|
||||
@@ -1294,6 +1438,10 @@ async def stream_llm(url: str, model: str, messages: List[Dict], temperature: fl
|
||||
model, messages_copy, temperature, max_tokens,
|
||||
stream=True, tools=tools, num_ctx=get_context_length(url, model),
|
||||
)
|
||||
elif provider == "chatgpt-subscription":
|
||||
target_url = _normalize_chatgpt_subscription_url(url)
|
||||
h = _provider_headers(provider, headers)
|
||||
payload = _build_chatgpt_responses_payload(model, messages_copy, temperature, max_tokens, stream=True)
|
||||
else:
|
||||
target_url = url
|
||||
payload = {
|
||||
@@ -1325,6 +1473,68 @@ async def stream_llm(url: str, model: str, messages: List[Dict], temperature: fl
|
||||
return
|
||||
note_model_activity(target_url, model)
|
||||
|
||||
# ── ChatGPT Subscription / Codex Responses streaming ──
|
||||
if provider == "chatgpt-subscription":
|
||||
event_name = ""
|
||||
input_tokens = 0
|
||||
output_tokens = 0
|
||||
try:
|
||||
client = _get_http_client()
|
||||
async with client.stream('POST', target_url, json=payload, headers=h, timeout=stream_timeout) as r:
|
||||
_clear_host_dead(target_url)
|
||||
if r.status_code != 200:
|
||||
raw = (await r.aread()).decode(errors="replace")
|
||||
friendly = _format_chatgpt_subscription_error(r.status_code, raw)
|
||||
yield f'event: error\ndata: {json.dumps({"status": r.status_code, "text": friendly, "raw": raw[:500]})}\n\n'
|
||||
return
|
||||
async for line in r.aiter_lines():
|
||||
if not line:
|
||||
continue
|
||||
if line.startswith("event:"):
|
||||
event_name = line[6:].strip()
|
||||
continue
|
||||
if not line.startswith("data:"):
|
||||
continue
|
||||
raw = line[5:].strip()
|
||||
if not raw:
|
||||
continue
|
||||
try:
|
||||
data = json.loads(raw)
|
||||
except json.JSONDecodeError:
|
||||
continue
|
||||
evt = data.get("type") or event_name
|
||||
if evt == "response.output_text.delta":
|
||||
delta = data.get("delta") or ""
|
||||
if delta:
|
||||
yield f'data: {json.dumps({"delta": delta})}\n\n'
|
||||
elif evt == "response.completed":
|
||||
usage = (data.get("response") or {}).get("usage") or data.get("usage") or {}
|
||||
input_tokens = usage.get("input_tokens") or usage.get("prompt_tokens") or input_tokens
|
||||
output_tokens = usage.get("output_tokens") or usage.get("completion_tokens") or output_tokens
|
||||
if input_tokens or output_tokens:
|
||||
yield f'data: {json.dumps({"type": "usage", "data": {"input_tokens": input_tokens, "output_tokens": output_tokens}})}\n\n'
|
||||
yield "data: [DONE]\n\n"
|
||||
return
|
||||
elif evt in ("response.failed", "error"):
|
||||
err = data.get("error") or (data.get("response") or {}).get("error") or {}
|
||||
text = err.get("message") if isinstance(err, dict) else str(err or "ChatGPT Subscription request failed")
|
||||
yield f'event: error\ndata: {json.dumps({"status": 502, "text": text})}\n\n'
|
||||
return
|
||||
yield "data: [DONE]\n\n"
|
||||
except (httpx.ConnectError, httpx.ConnectTimeout) as e:
|
||||
_cooled = _mark_host_dead(target_url)
|
||||
_tail = f" — host cooled for {DEAD_HOST_COOLDOWN:.0f}s" if _cooled else " — transient, will retry"
|
||||
logger.warning(f"ChatGPT Subscription stream connect to {target_url} failed: {e}{_tail}")
|
||||
yield f'event: error\ndata: {json.dumps({"error": f"Cannot reach {_host_key(target_url)}", "status": 503})}\n\n'
|
||||
except httpx.ReadTimeout:
|
||||
yield f'event: error\ndata: {json.dumps({"error": "Read timeout", "status": 504})}\n\n'
|
||||
except httpx.NetworkError:
|
||||
yield f'event: error\ndata: {json.dumps({"error": "Network error", "status": 502})}\n\n'
|
||||
except Exception as e:
|
||||
logger.error(f"ChatGPT Subscription stream error: {e}")
|
||||
yield f'event: error\ndata: {json.dumps({"error": str(e), "status": 502})}\n\n'
|
||||
return
|
||||
|
||||
# ── Native Ollama streaming ──
|
||||
if provider == "ollama":
|
||||
_ollama_tool_calls: List[Dict] = []
|
||||
|
||||
Reference in New Issue
Block a user