Back to Prompt Library
planning
Initial Agent Design for Code Generation
Inspect the original prompt language first, then copy or adapt it once you know how it fits your workflow.
Linked challenge: Self-Improving GPT-5.3-Codex Agent for Code Generation & Refinement
Format
Code-aware
Lines
1
Sections
1
Linked challenge
Self-Improving GPT-5.3-Codex Agent for Code Generation & Refinement
Prompt source
Original prompt text with formatting preserved for inspection.
1 lines
1 sections
No variables
0 checklist items
Outline the architecture of your self-improving code generation agent using the OpenAI Agents SDK. Define the primary agent's role (e.g., 'Code Master') and its interaction loop, which should include steps like: `Understand Problem -> Generate Code -> Execute Tests -> Analyze Feedback -> Refine Code`. Describe how GPT-5.3-Codex will be leveraged in each of these steps, especially for generating initial code and then for debugging based on test feedback. Provide a conceptual Python class or function structure for your agent using `openai.beta.assistants`.
Adaptation plan
Keep the source stable, then change the prompt in a predictable order so the next run is easier to evaluate.
Keep stable
Preserve the role framing, objective, and reporting structure so comparison runs stay coherent.
Tune next
Swap in your own domain constraints, anomaly thresholds, and examples before you branch variants.
Verify after
Check whether the prompt asks for the right evidence, confidence signal, and escalation path.