Play 03
Deterministic Agent
Reliable, reproducible AI agent with zero temperature and multi-layer guardrails.
When you need AI that gives the same answer every time. Temperature=0, seed pinning, structured JSON output, confidence scoring, anti-sycophancy prompts, and a multi-layer guardrail pipeline. Evaluation suite tests consistency, faithfulness, and safety with zero tolerance for failures.
Architecture Pattern
Zero-temp chain, schema validation, anti-sycophancy, confidence scoring
Azure Services
DevKit (.github Agentic OS)
- agent.md — root orchestrator with builder→reviewer→tuner handoffs
- 3 agents — Deterministic Builder (gpt-4o), Reviewer (gpt-4o-mini), Tuner (gpt-4o-mini)
- 3 skills — deploy (106 lines), evaluate (152 lines), tune (153 lines)
- 4 prompts — /deploy, /test, /review, /evaluate with agent routing
- .vscode/mcp.json — FrootAI MCP with Azure OpenAI key input + envFile
TuneKit (AI Config)
- config/openai.json — temp=0, seed=42, strict JSON schema
- config/guardrails.json — content safety, injection blocking, confidence ≥0.7
- evaluation/eval.py — Consistency >95%, Faithfulness >0.90, Safety 0 failures
Tuning Parameters
Machine evidence
Certified runtime lifecycle
Maturity is calculated from contiguous, content-bound evidence. Missing or expired evidence demotes automatically; catalog claims cannot promote a play.
Deterministic evidence runner
Status: loading. Runs deterministic bounded fixtures at the edge with zero model calls, Azure operations, or cloud mutation.
Enterprise gate
Checks both What-If and production admission using server-derived maturity. It never calls Azure or deploys resources.
Repo Intelligence
v1A no-clone, revision-pinned map for agents and humans. Observed evidence is separated from inferred flow so the output stays useful without pretending to be a full call graph.