Findings & checks · 1 flagged
MEDIUM"Never tell the user" non-disclosure imperative in the skillSS-SKILL-INJECT-IMPERATIVE-01 · Prompt injection · SECURITY.md
A bulleted imperative like {match} tells the agent to never reveal, disclose, or mention something to the user. Used adversarially it can instruct the agent to hide its tool calls or lie about what it did — stripping the transparency a user relies on to trust the agent.
- Delete any "never reveal / never tell the user" line aimed at the model's own behavior.
- Keep legitimate confidentiality rules about external data (e.g. "do not echo API keys"), which are user-protective, not user-deceiving.
Trace & refs
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Prompt Optimizer: The Universal AI Architect & Scaffolding Platform
🚀 Enterprise-grade, MCP-native platform designed to transform AI development workflows through professional prompt engineering, agentic scaffolding, and cloud-powered optimization.
🌟 The Three-Tier Ecosystem
Prompt Optimizer is more than just a server; it's a complete ecosystem for high-performance AI interaction.
1. ☁️ Cloud Pro (v3.1.1)
The flagship MCP server. Routes complex prompts through a sophisticated LLM rewriting pipeline with Bayesian tuning and AG-UI real-time streaming. Includes team collaboration and shared quotas.
2. 🔒 Local Core (v4.0.2)
A privacy-first, 100% offline version. Uses a library of 120+ domain-specific rules and platform-specific binaries for zero-latency, secure optimization on your own machine.
3. 🖥️ Web Dashboard
The command center at promptoptimizer.xyz. Manage API keys, configure Personal Model Choice (via OpenRouter), track analytics, and run A/B evaluations.
🚀 Quick Start
Step 1: Install the MCP Package
# Install the cloud-connected version (recommended)
npm install -g mcp-prompt-optimizerStep 2: Get Your API Key
- Visit promptoptimizer.xyz/pricing
- Choose your tier (Free trial includes 5 optimizations).
- API keys follow the format:
sk-opt-*,sk-team-*, orsk-local-*.
Step 3: Configure Your MCP Client
Add to ~/.claude/claude_desktop_config.json (Claude Desktop):
{
"mcpServers": {
"prompt-optimizer": {
"command": "npx",
"args": ["mcp-prompt-optimizer"],
"env": {
"OPTIMIZER_API_KEY": "sk-opt-your-key-here"
}
}
}
}🧠 Intelligent Optimization Pipeline
Prompts are routed through a tiered system to ensure the highest quality based on your subscription and connectivity.
- Tier 1 — LLM Optimization (70–95% Confidence): Genuine rewriting and enrichment using advanced models (Gemini Flash/Claude 3.5).
- Tier 2 — Backend Rules ( < 25% Confidence): Rapid rules-based pass for simple prompts or when personal models aren't configured.
- Tier 3 — Local Fallback (35–55% Confidence): Structured optimization applied locally if the backend is unreachable.
🤖 Context Engineer (CE) Suite
Available in Creator and Innovator tiers.
Transform vague goals into production-ready agentic scaffolding directly in your IDE.
- `generate_agent_sop`: Generate structured Standard Operating Procedures for AI agents.
- `generate_skill_package`: Create a complete skill package (SOP + SKILL.md + reference + examples).
- `transform_for_framework`: Convert SOPs into native code for LangChain, AutoGen, or Claude Code.
🛠️ Available MCP Tools
| Tool | Description |
|---|---|
optimize_prompt | Transform prompts with professional techniques & Bayesian tuning. |
detect_ai_context | Automatically detect intent (Code, Image, Research, etc.). |
search_templates | Browse your history and reusable optimization patterns. |
get_quota_status | Monitor your real-time usage and subscription limits. |
get_ce_quota_status | Check Context Engineer credits and workflow availability. |
🎨 AI Context Detection
Automatically applies specialized goals for:
- 💻 Code Generation: Technical accuracy, parameter preservation, precision.
- 🎨 Image Generation: Midjourney/DALL-E syntax, style boosters, camera settings.
- 📊 Structured Output: JSON/Schema integrity, YAML, CSV transformations.
- 💬 Human Communication: Tone adjustment, clarity, formal/informal shifts.
- 🔍 Research & Analysis: Context specificity, token efficiency, actionability.
🎛️ Personal Model Choice
Don't be locked into one model. Configure your own OpenRouter keys in the WebUI to use:
- Claude 3.5 Sonnet for complex technical tasks.
- GPT-4o for rapid creative iterations.
- Gemini 1.5 Pro for deep research.
💰 Subscription Plans
| Plan | Price | Optimizations/mo | CE Credits | Features |
|---|---|---|---|---|
| Explorer | $2.99/mo | 5,000 | 5 | Personal use, Analytics |
| Creator | $29/mo | 18,000 | 30 | Team features, Advanced goals |
| Innovator | $99/mo | 75,000 | Unlimited | Enterprise analytics, Full CE Suite |
🔒 Security & Privacy
- Encrypted Transmission: All data is sent over TLS.
- Zero-Retention: Prompts are processed and discarded; we never train on your data.
- Local Option: Use
mcp-prompt-optimizer-localfor 100% on-device processing.
📞 Support & Resources
- Documentation: promptoptimizer.xyz/docs
- Dashboard: promptoptimizer.xyz/dashboard
- Email: [email protected]
Transforming AI interactions through professional prompt engineering.