continuedev/continue · error · Error

no query params found

Error message

no query params found

What it means

Streaming-path model routing fallback: the model string maps to no known publisher, so there is no streaming implementation to invoke. Mirror of the non-streaming Unsupported model error.

Source

Thrown at core/context/mcp/MCPOauth.ts:42

const authenticatingContexts = new Map<string, MCPOauthContext>();

// Map state parameters to server URLs for OAuth callback matching
const stateToServerUrl = new Map<string, string>();

const PORT = 3000;

let serverInstance: http.Server | null = null;

const createServerForOAuth = () =>
  http.createServer((req, res) => {
    try {
      if (!req.url) {
        throw new Error("no url found");
      }

      const parsedUrl = url.parse(req.url, true);
      if (!parsedUrl.query["code"]) {
        throw new Error("no query params found");
      }

      const code = parsedUrl.query["code"] as string;
      const state = parsedUrl.query["state"] as string | undefined;

      void handleMCPOauthCode(code, state);

      const html = `
<!DOCTYPE html>
<html>
<head><title>Authentication Complete</title></head>
<body>Authentication Complete. You can close this page now.</body>
</html>`;

      res.writeHead(200, {
        "Content-Type": "text/html",
      });
      res.end(html);

View on GitHub (pinned to 5522c6f44c)

Solutions

  1. Use a supported model id (gemini-*, claude-*@*, mistral/codestral)
  2. Update openai-adapters to pick up new provider mappings
  3. Log body.model at call sites to catch typos early

Example fix

// before
api.chatCompletionStream({ model: 'llama-3.1-405b', ... }, signal);

// after
api.chatCompletionStream({ model: 'gemini-1.5-flash', ... }, signal);
Defensive patterns

Strategy: validation

Validate before calling

const routable = (m: string) => /^(gemini|claude|mistral|codestral)/.test(m); if (!routable(body.model)) throw new Error('unsupported model for Vertex streaming');

Type guard

const isVertexModel = (m: string): boolean => /^(gemini|claude|mistral|codestral)/.test(m);

Try / catch

try { ... } catch (e) { if ((e as Error).message.startsWith('Unsupported model')) { /* correct the model id */ } throw e; }

Prevention

When it happens

Trigger: Streaming a model id that determineVertexProvider can't classify — wrong prefix, typo, or unsupported family.

Common situations: Switching a streaming pipeline to a model name copied from another provider; adapter version predating support for a newly added Vertex model.

Related errors

AI-assisted analysis of continuedev/continue@5522c6f44c (2026-08-27). Data as JSON: /api/errors/90b5eee1fb6716fd. Report an issue: GitHub.