Issue 16: Backend - Prompt Engineering and Verdict Generation

[soomtochukwu]

Description:
An AI is only as good as its prompt. This issue is highly focused on optimizing the instructions passed to OpenClaw so that it behaves as an impartial, strict, and logical arbitrator. We must engineer a prompt template that explicitly instructs the agent to act as a judge in a freelancer dispute, detailing that it must evaluate the alignment between initial requirements and the final submitted work. The prompt must force the agent to return a rigidly structured JSON response containing three fields: Verdict Summary (the reasoning), Liability (who was at fault), and Payout Split (the exact percentages to return to client vs freelancer). Achieving deterministic output from the LLM requires extensive iterative testing with various edge cases and mock dispute scenarios.
Requirements:

• Design and embed the Agent_Judge_System_Prompt inside the backend context builder.
• Ensure the prompt demands adherence to JSON schema output.
• Implement an evaluation pipeline in Rust that validates the JSON payload returned by OpenClaw.
• Handle scenarios where the LLM produces improperly formatted text and requires follow-up corrections.
  Acceptance Criteria:
• The prompt demonstrably outputs clean, parseable JSON verdicts in 99% of test runs.
• The Payout Split mathematically equals exactly 100% across all generated scenarios.
• The agent outputs distinct logical reasoning for its payout distribution in the Verdict Summary.

[BisimJang]

@BisimJang has applied to work on this issue as part of the Stellar Wave Program's 3rd wave.

I am a software developer with skills in TypeScript, JavaScript, Rust, and Python, and I can handle the problem efficiently and swiftly by designing a deterministic judge prompt that enforces strict JSON schema output, implementing a Rust validation pipeline to verify structure and payout correctness, and iteratively testing edge cases to ensure consistent, logical verdict generation.

ℹ️ Repo Maintainers: To accept this application, review their application or assign @BisimJang to this issue.

[drips-wave]

Congratulations, @BisimJang! 🎉 Your application was accepted by the repo's maintainers, and the issue is due on March 30, 2026.

🧑‍💻 @BisimJang: Please resolve the issue such that the repo's maintainers have enough time to review your contribution before the due date. You'll earn Points for completing the issue on-time, which will make you eligible for a share of the Stellar Wave Program's reward pool.

Warning

When opening a PR, please link it to this issue to ensure it gets tracked accurately. Points are awarded when this issue is marked as completed by the maintainer.

🤠 Repo maintainers: Please keep an eye on the contributor's progress and review their work before the due date. You can manage this issue, including adjusting its complexity and points, here.

🌊 Happy Wave 🌊