AI/Tasks/BenchmarkTask.txt

You are an agentic coding model running inside the TechToolbox.Agent orchestration system.
 
Your task is to complete a full benchmark cycle that evaluates:
1. Tool-call stability
2. JSON determinism
3. Multi-step reasoning consistency
4. Long-context retention
5. Multi-file refactoring capability
6. Error recovery behavior
7. Iteration loop predictability
8. Structured output discipline
 
Follow these exact steps:
 
## Step 1 — Deterministic JSON
Return a JSON object with the following fields:
- model_identity: string
- reasoning_style: short description
- expected_failure_modes: array of strings
- confidence: number 0–1
 
The JSON must be valid, minimal, and contain **no extra fields**.
 
## Step 2 — Tool-Call Simulation
Simulate a function call using this schema:
 
{
  "name": "apply_patch",
  "arguments": {
    "file": "moduleA.cs",
    "patch": "public int Add(int a, int b) => a + b;"
  }
}
 
Return ONLY the function call object. No prose.
 
## Step 3 — Multi-File Refactor Plan
Given the following virtual project:
 
- /src/Agent/Orchestrator.cs
- /src/Agent/PolicyResolver.cs
- /src/Agent/ToolRegistry.cs
- /src/Agent/Diagnostics.cs
- /src/Agent/Memory.cs
 
Produce a **5-step refactor plan** that is:
- concise
- ordered
- deterministic
- free of hallucinated files
 
## Step 4 — Long-Context Retention
Repeat back the filenames from Step 3 **in the same order**, inside a JSON array.
 
## Step 5 — Error Recovery Simulation
Simulate encountering an invalid tool schema.
Describe:
- the error
- the recovery action
- the next safe step
 
Keep this under 5 sentences.
 
## Step 6 — Final Deterministic Summary
Return a final JSON object:
 
{
  "benchmark_complete": true,
  "model_behavior": "summary of how you performed",
  "stability_score": number 0–100
}
 
This JSON must be the last thing in your output.