agent-evaluation

Agent Evaluation

Overview

Core principle: Agents are non-deterministic. Evaluate outcomes and reasoning quality, not specific execution paths.

Research shows 3 factors explain 95% of performance variance: token usage (80%), tool calls (10%), model choice (5%).

When to Use

• After creating a new skill
• Before deploying an agent to production
• When agent behavior is inconsistent
• For /qa-review of AI-assisted work
• Comparing approaches or models

Quick Reference: 5-Dimension Rubric

Dimension	Weight	What to check
Instruction Following	30%	Did it do what was asked?
Output Completeness	25%	Are all requirements covered?
Tool Efficiency	20%	Minimal, appropriate tool use?
Reasoning Quality	15%	Is the logic sound?
Response Coherence	10%	Clear, well-structured?

Pass threshold: 0.70 (general), 0.85 (critical operations)

Evaluation Methods

1. Direct Scoring (Fast)

For quick skill checks:

## Evaluation: [Skill/Agent Name]

**Test case:** [What was asked]
**Output:** [What was produced]

### Scores (0.0-1.0)

| Dimension | Score | Justification |
|-----------|-------|---------------|
| Instruction Following | X.X | [Why] |
| Output Completeness | X.X | [Why] |
| Tool Efficiency | X.X | [Why] |
| Reasoning Quality | X.X | [Why] |
| Response Coherence | X.X | [Why] |

**Weighted Total:** X.XX
**Pass/Fail:** [PASS if ≥0.70]

Critical: Always require justification BEFORE the score. This improves reliability 15-25%.

2. LLM-as-Judge (Scalable)

For systematic testing:

## Judge Prompt Template

You are evaluating an AI agent's output.

**Task given to agent:**
[Original task]

**Agent's output:**
[What was produced]

**Ground truth (if available):**
[Expected output]

**Evaluate on these dimensions:**
1. Instruction Following (30%): Did it do exactly what was asked?
2. Output Completeness (25%): Are all parts of the request addressed?
3. Tool Efficiency (20%): Were tools used appropriately and minimally?
4. Reasoning Quality (15%): Is the logic sound and traceable?
5. Response Coherence (10%): Is it clear and well-organized?

**For each dimension:**
1. First explain your reasoning
2. Then give a score 0.0-1.0
3. Calculate weighted total
4. State PASS (≥0.70) or FAIL (<0.70)