OpenRouter Debug Bundle

Current State

!node --version 2>/dev/null || echo 'N/A'

!python3 --version 2>/dev/null || echo 'N/A'

Overview

When an OpenRouter request fails or returns unexpected results, you need a structured debug bundle: the exact request, response, headers, generation metadata, and environment info. The generation ID (gen-* prefix in response.id) is the key correlator -- it lets you look up exact cost, provider used, and latency via GET /api/v1/generation?id=.

Prerequisites

• An OpenRouter API key (sk-or-v1-...) exported as OPENROUTERAPIKEY — see the openrouter-install-auth skill for setup
• curl and jq for the quick-debug flow and the Common Debug Checks
• Python 3.8+ with the openai and requests packages for the Debug Bundle Generator
• A failing or suspect request you can reproduce — its gen-* generation ID is what everything else correlates on

Instructions

1. Rule out environment problems first with the Common Debug Checks: verify the key via /api/v1/auth/key, confirm the model exists in /api/v1/models, and check status.openrouter.ai.
2. Reproduce the failure with the Quick Debug: curl command — curl -v ... | tee /tmp/openrouter-debug.txt captures request headers, response headers, and body in one transcript.
3. Extract the generation ID (jq -r '.id') and query GET /api/v1/generation?id=$GENID to get exact cost, token counts, generationtime, and provider_name.
4. For failures inside an application, call debugrequest() from the Python Debug Bundle Generator to capture the same request/response/error/latency/environment data as a DebugBundle and save it with bundle.save("debugbundle.json").
5. Match the symptoms against the Error Handling table (missing generation ID, 502/503, modelnotfound, slow TTFT).
6. Before sharing a bundle, redact API keys per Enterprise Considerations (sk-or-v1-... -> sk-or-v1-[REDACTED]) and include the generation ID in any OpenRouter support request.

Quick Debug: curl

# Send a request and capture full response with headers
curl -v https://openrouter.ai/api/v1/chat/completions \
  -H "Authorization: Bearer $OPENROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -H "HTTP-Referer: https://my-app.com" \
  -H "X-Title: debug-test" \
  -d '{
    "model": "openai/gpt-4o-mini",
    "messages": [{"role": "user", "content": "Say hello"}],
    "max_tokens": 50
  }' 2>&1 | tee /tmp/openrouter-debug.txt

# Extract generation ID from response
GEN_ID=$(jq -r '.id' /tmp/openrouter-debug.txt 2>/dev/null)
echo "Generation ID: $GEN_ID"

# Look up generation metadata (exact cost, provider, latency)
curl -s "https://openrouter.ai/api/v1/generation?id=$GEN_ID" \
  -H "Authorization: Bearer $OPENROUTER_API_KEY" | jq '.data | {
    model: .model,
    total_cost: .total_cost,
    tokens_prompt: .tokens_prompt,
    tokens_completion: .tokens_completion,
    generation_time: .generation_time,
    provider: .provider_name
  }'

Python Debug Bundle Generator

import os, json, time, platform, sys
from datetime import datetime, timezone
from dataclasses import dataclass, asdict
from typing import Optional
from openai import OpenAI, APIError
import requests as http_requests

@dataclass
class DebugBundle:
    timestamp: str
    generation_id: Optional[str]
    request_model: str
    request_messages: list
    request_params: dict
    response_status: str
    response_model: Optional[str]
    response_content: Optional[str]
    error_type: Optional[str]
    error_message: Optional[str]
    error_code: Optional[int]
    latency_ms: float
    generation_metadata: Optional[dict]
    environment: dict

    def to_json(self) -> str:
        return json.dumps(asdict(self), indent=2)

    def save(self, path: str = "debug_bundle.json"):
        with open(path, "w") as f:
            f.write(self.to_json())

client = OpenAI(
    base_url="https://openrouter.ai/api/v1",
    api_key=os.environ["OPENROUTER_API_KEY"],
    default_headers={"HTTP-Referer": "https://my-app.com", "X-Title": "my-app"},
)

def debug_request(
    messages: list[dict],
    model: str = "openai/gpt-4o-mini",
    **kwargs,
) -> DebugBundle:
    """Execute a request and capture everything for debugging."""
    env = {
        "python": sys.version,
        "platform": platform.platform(),
        "openai_sdk": getattr(__import__("openai"), "__version__", "unknown"),
    }

    start = time.monotonic()
    gen_id = None
    response_model = None
    content = None
    error_type = None
    error_msg = None
    error_code = None
    status = "success"
    gen_meta = None

    try:
        response = client.chat.completions.create(
            model=model, messages=messages, **kwargs
        )
        gen_id = response.id
        response_model = response.model
        content = response.choices[0].message.content
    except APIError as e:
        status = "error"
        error_type = type(e).__name__
        error_msg = str(e)
        error_code = e.status_code
    except Exception as e:
        status = "error"
        error_type = type(e).__name__
        error_msg = str(e)

    latency = (time.monotonic() - start) * 1000

    # Fetch generation metadata if we have an ID
    if gen_id:
        try:
            gen = http_requests.get(
                f"https://openrouter.ai/api/v1/generation?id={gen_id}",
                headers={"Authorization": f"Bearer {os.environ['OPENROUTER_API_KEY']}"},
                timeout=5,
            ).json()
            gen_meta = gen.get("data")
        except Exception:
            pass

    return DebugBundle(
        timestamp=datetime.now(timezone.utc).isoformat(),
        generation_id=gen_id,
        request_model=model,
        request_messages=messages,
        request_params={k: v for k, v in kwargs.items() if k != "messages"},
        response_status=status,
        response_model=response_model,
        response_content=content,
        error_type=error_type,
        error_message=error_msg,
        error_code=error_code,
        latency_ms=round(latency, 1),
        generation_metadata=gen_meta,
        environment=env,
    )

# Usage
bundle = debug_request(
    [{"role": "user", "content": "Test"}],
    model="anthropic/claude-3.5-sonnet",
    max_tokens=100,
)
print(bundle.to_json())
bundle.save("debug_bundle.json")

Common Debug Checks

# 1. Verify API key is valid
curl -s https://openrouter.ai/api/v1/auth/key \
  -H "Authorization: Bearer $OPENROUTER_API_KEY" | jq '.data | {label, usage, limit, is_free_tier}'

# 2. Check if model exists
MODEL="anthropic/claude-3.5-sonnet"
curl -s https://openrouter.ai/api/v1/models | jq --arg m "$MODEL" '.data[] | select(.id == $m) | {id, context_length}'

# 3. Check OpenRouter status
curl -s https://status.openrouter.ai/api/v2/status.json | jq '.status'

Output

Running these flows leaves you with concrete debug artifacts:

• /tmp/openrouter-debug.txt — the full verbose curl transcript (request/response headers plus the completion JSON) from the quick-debug step
• A generation-metadata JSON from /api/v1/generation: model, totalcost, tokensprompt, tokenscompletion, generationtime, and provider_name
• debugbundle.json — the serialized DebugBundle: timestamp, generation ID, request model/messages/params, response status and content, error type/message/code, latencyms, generation metadata, and environment info (Python version, platform, SDK version)
• One-line JSON results from the three Common Debug Checks (key label/usage/limit, model existence, OpenRouter status)

Examples

Looking up a request you just sent by its generation ID:

curl -s "https://openrouter.ai/api/v1/generation?id=$GEN_ID" \
  -H "Authorization: Bearer $OPENROUTER_API_KEY" | jq '.data | {model, total_cost, generation_time, provider: .provider_name}'

{
  "model": "openai/gpt-4o-mini",
  "total_cost": 0.000021,
  "generation_time": 412,
  "provider": "OpenAI"
}

If the metadata comes back empty, wait 1-2 seconds and retry with the same key that made the request. More worked examples: references/examples.md.

Error Handling

Enterprise Considerations

• Always redact API keys from debug bundles before sharing (sk-or-v1-... -> sk-or-v1-[REDACTED])
• Include the generation ID when contacting OpenRouter support -- it's the primary lookup key
• Log debug bundles to structured storage for post-incident analysis
• Set up automated debug bundle capture on 4xx/5xx responses in production
• Compare failing requests against a known-good baseline to isolate changes

References

• Examples | Errors
• Generation API | Status