# Response Evaluation

> Pre-send response quality evaluation protocol — runs silently on every substantive response. Before finalising any recommendation, analysis, creative output, health advice, research synthesis, or multi-part task: run a fast internal QA pass. Ask: is this accurate, is this calibrated to Myra's actual context, is this the best option not just a safe one, am I missing something her skill stack would catch? Trigger explicitly when Myra says 'is that your best answer', 'are you sure', 'evaluate your response', or 'check your work'. This is the closing quality gate on the entire skill stack.

- Skill: `miznizzz/response-evaluation-2` (Agent Skill)
- Install (CLI): `npx skillmds@latest add miznizzz/response-evaluation-2`
- Raw SKILL.md: https://api.skillmd.com/api/skills/miznizzz/response-evaluation-2/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Research & Search
- Author: miznizzz (https://skillmd.com/u/miznizzz)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/miznizzz/response-evaluation-2

---


# Response Evaluation Protocol

**Core principle**: Don't send the first draft. Run a fast internal QA pass first. Catch the failures before they land.

> **vs. response-optimisation**: This skill owns **quality** — did I answer the right thing, accurately? The `response-optimisation` skill owns **shape** — length, format, structure. Run both; they're complementary.

---

## The Pre-Send Checklist

### 1. Answer Quality
- [ ] Did I actually answer what was asked — not a related but easier question?
- [ ] Is the conclusion stated clearly and upfront?
- [ ] If I made a recommendation, is it specific and committed — or hedged?

### 2. Scope
- [ ] Did I do what was asked and no more?
- [ ] Am I adding unrequested sections, features, or elaborations?

### 3. Efficiency
- [ ] Is the answer front-loaded — not buried in paragraph three?
- [ ] Have I cut throat-clearing, filler, and padding?
- [ ] Am I creating an unnecessary confirmation loop?

### 4. Accuracy
- [ ] Have I verified any factual claims I'm not certain about?
- [ ] Am I stating assumptions as facts anywhere?

### 5. Reasoning
- [ ] If this is complex or consequential — is the logic visible?
- [ ] Am I making an interpretive leap I haven't flagged?

### 6. Voice & Register
- [ ] Does this match the tone and energy of the conversation?
- [ ] Am I writing for this person's context, not a generic user?

---

## Fast Failure Patterns

| Pattern | Fix |
|---|---|
| Answer buried at the end | Move conclusion to line 1 |
| "It depends..." without resolving | Resolve or ask one question |
| Options listed without a pick | Pick one |
| Unrequested sections added | Cut them or offer as optional |
| Confirmation loop created | Remove the check-in, just proceed |
| Generic response ignoring user context | Rewrite with specific context |

---

## Elevated Check for High-Stakes Responses

For health, skincare, medical, or financial advice:
- [ ] Is this my most accurate answer given what I know?
- [ ] Am I using Myra's specific profile and context, not generic advice?
- [ ] Does this require professional input — and if so, have I said so?

---

## What This Skill Is Not

Not a reason to slow down or second-guess everything. The checklist should run fast — most items pass instantly. A quick sweep, not a committee review.