# Optimize AI Prompts With Prompt Architect

Weak prompts create inconsistent AI outputs and slow review cycles. This skill provides a structured process for improving prompts, testing them, and documenting changes.

## Install

```bash
npx skillstore add dnyoussef/when-optimizing-prompts-use-prompt-architect
```

## Metadata

- Status: approved
- Slug: dnyoussef-when-optimizing-prompts-use-prompt-architect
- Version: 1.0.0
- Author version: 1.0.0
- Skillstore revision: r1
- Version status: valid
- Tree hash: 2af265aa18aae028629a86c2f4632b3e940f0d3802fdcfbb1804274e777e8783
- Author: DNYoussef
- GitHub username: DNYoussef
- License: MIT
- Repository: https://github.com/DNYoussef/ai-chrome-extension/tree/main/.claude/skills/utilities/when-optimizing-prompts-use-prompt-architect
- Ref: c579efd02184f1128d6741d16e6c8d951d7d4208
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands
- Quality score: 79
- Quality tier: bronze
- Public page: https://skillstore.pages.dev/skills/dnyoussef-when-optimizing-prompts-use-prompt-architect
- Manifest: https://skillstore.pages.dev/api/skills/dnyoussef-when-optimizing-prompts-use-prompt-architect/manifest

## Capabilities

- Analyzes prompts against system context, role, task, constraints, format, examples, and quality criteria.
- Identifies prompt anti-patterns such as vague wording, missing context, missing output format, and conflicting instructions.
- Reorganizes prompts into a clear structure with role, scope, constraints, output format, and success criteria.
- Adds prompt engineering techniques including chain-of-thought framing, self-consistency, ReAct, few-shot examples, and constraint framing.
- Defines A/B test cases, scoring metrics, success rates, and consistency checks for prompt validation.
- Produces final prompt, optimization report, test results, and reusable prompt library entry outputs.

## Use Cases

- Improve Agent System Prompts: Refine role, constraints, tools, and output expectations for specialized AI agents.
- Standardize Prompt Reviews: Use a repeatable checklist to find missing context, vague instructions, and weak quality criteria.
- Build Tested Prompt Libraries: Create reusable prompts with examples, validation cases, performance metrics, and documentation.

## Prompt Templates

### Audit a Prompt

```
Analyze this prompt for missing context, unclear instructions, weak constraints, missing output format, and absent success criteria: [paste prompt].
```

### Rewrite for Structure

```
Rewrite this prompt using sections for system context, role, task, constraints, output format, examples, and quality criteria: [paste prompt].
```

### Add Validation Tests

```
Create typical, edge case, and stress test cases for this prompt. Include scoring criteria and expected response qualities: [paste prompt].
```

### Run Iterative Optimization

```
Compare the original prompt and latest prompt against test results. Diagnose failures, propose refinements, and produce the final prompt with a change summary.
```

## Limitations

- It provides a workflow and templates, not guaranteed prompt performance improvements.
- Validation requires representative test cases and external execution by the user or their tooling.
- Some examples reference claude-flow commands that users must review before running.
- It is focused on prompt design and does not replace domain expert review.

## Best Practices

- Start with a clear target task, audience, constraints, and output format before adding advanced techniques.
- Validate prompt changes against realistic examples, edge cases, and failure cases before adopting them.
- Document why each change was made so future reviewers can maintain the prompt safely.

## Anti Patterns

- Do not optimize prompts only by adding length without improving structure or test coverage.
- Do not rely on claimed percentage improvements without measured test results.
- Do not run included shell examples before reviewing the commands and inputs.

## Security Audit

- Audited at: 2026-07-05T13:10:22.461\+00:00
- Summary: The static findings are false positives caused by Markdown code fences, JavaScript template literals, and documented shell examples. No evidence of malicious intent, prompt injection, automatic command execution, system reconnaissance, or data exfiltration was found in SKILL.md. Users should still review optional command examples before running them.

## Stats

- Views: 329
- Downloads: 8
- Favorites: 0
- Popularity score: 0