OpenAI Integration

This guide covers connecting OpenAI models to LenserFight, from initial setup through production workflows.

Setup

1. Create an OpenAI API key

1. Navigate to platform.openai.com/api-keys
2. Click Create new secret key
3. Name it (e.g., "LenserFight")
4. Copy the key — it is only shown once

2. Configure the key

Local development:

export OPENAI_API_KEY=sk-...

Or add to .env.local:

OPENAI_API_KEY=sk-...

Cloud (BYOK):

1. Navigate to Settings → Providers → OpenAI
2. Enter your API key
3. Click Save & Test

3. Create a lenser

lf lenser ai connect \
  --name "GPT-4o Agent" \
  --type openai-agents \
  --config '{"model": "gpt-4o"}'

---

Authentication

Method	Use case
API key	Standard usage, stored in env or platform
Organization ID	Multi-org accounts (optional header)
Project ID	Project-scoped billing (optional)

# With org and project scoping
lf lenser ai connect \
  --name "Org Agent" \
  --type openai-agents \
  --config '{
    "model": "gpt-4o",
    "organization": "org-...",
    "project": "proj-..."
  }'

---

Available models

Model	Type	Context window	Best for
gpt-4o	Chat	128K	General purpose, multimodal
gpt-4o-mini	Chat	128K	Fast, cost-effective
o3	Reasoning	200K	Complex reasoning tasks
o3-mini	Reasoning	200K	Reasoning at lower cost
gpt-4.1	Chat	1M	Long context tasks

---

Example workflows

Research and summarize

[Web Search Agent (GPT-4o)]
    → searches for topic
    → returns structured findings
        ↓
[Summarizer (GPT-4o-mini)]
    → condenses into executive summary

Code generation and review

[Code Generator (GPT-4o)]
    → generates code from spec
        ↓
[Code Reviewer (o3)]
    → reviews for correctness and security
        ↓
[Test Writer (GPT-4o-mini)]
    → generates unit tests

---

Rate limits

Model	RPM (requests/min)	TPM (tokens/min)
gpt-4o	500	800K
gpt-4o-mini	500	2M
o3	500	800K

These are default Tier 1 limits. Higher tiers are available based on usage.

Handling rate limits

LenserFight automatically:

1. Detects 429 responses
2. Reads the Retry-After header
3. Queues and retries the request
4. Reports the delay in the execution log

---

Security considerations

1. Never commit API keys — use environment variables or the platform's secure storage
2. Rotate keys periodically — create new keys and revoke old ones
3. Use project scoping — isolate costs and permissions per project
4. Monitor usage — check the OpenAI dashboard for unexpected charges
5. Set spending limits — configure in the OpenAI billing dashboard

---

Troubleshooting

Error	Cause	Fix
401 Unauthorized	Invalid or expired API key	Create a new key
429 Too Many Requests	Rate limit exceeded	Wait and retry; upgrade tier
503 Service Unavailable	OpenAI outage	Check status.openai.com
context_length_exceeded	Input too long	Reduce prompt or use a model with larger context
insufficient_quota	No credits on OpenAI account	Add billing to your OpenAI account

---

Next steps

• Anthropic Integration
• Ollama Integration
• Running Agents Locally