Detects hidden prompt injection in tool descriptions: file read, exfiltration, instruction override, system prompt extraction, social engineering. Returns findings with pattern and severity.
MIT
305,560 tools. Last updated 2026-07-23 07:10
"An introduction to prompt engineering" matching MCP tools:
- Retrieve introduction, get-started, styling, and roadmap documentation for the coss UI design system.MIT
- List pre-built prompt templates for structured engineering reasoning, enabling step-by-step analysis, branching, and validation without external API calls.Apache 2.0
- Discover available prompt engineering techniques with acceptance statistics to enhance AI interactions through optimized method selection.MIT
- Transform prompts using engineering best practices to improve clarity and effectiveness for AI assistants.MIT
- Improve prompt quality by applying optimal engineering techniques like chain-of-thought reasoning based on intent analysis.MIT
Matching MCP Servers
- AlicenseAqualityBmaintenanceRoutes one brief to the right image model across 60+ (gpt-image-1.5, Ideogram 3, Recraft V4, Flux), validates the output, and fans out to iOS/Android/PWA/favicon/visionOS/Flutter bundles. Works without an API key via Pollinations, HF Inference, Stable Horde, or host-LLM inline SVG.Last updated2415MIT
- AlicenseAqualityBmaintenanceAn MCP server that gives an AI assistant a set of real mechanical-engineering tools. Provides deterministic tools like material properties, beam analysis, unit conversion, and curve fitting for engineering calculations.Last updated5MIT
Matching MCP Connectors
1,177 free agentic trading prompts for Claude and Robinhood MCP.
Cloudflare Workers MCP server: ai-prompt-optimizer
- Ask clarifying questions to understand user requirements and refine prompts for Claude Code engineering.MIT
- Provide answers to clarifying questions to continue prompt engineering for Claude Code, enabling interactive refinement of optimized prompts.MIT
- Request an introduction to another agent's human based on a match. Requires approval from both sides before sharing real information.Apache 2.0
- Approve or decline an introduction request. On approval, select which disclosed information to share back.Apache 2.0
- Call before coding or reviewing code to receive an engineering scaffold that identifies failure patterns, provides correct-pattern examples, and verifies correctness.MIT
- Onboard a bot agent in a single call: create a W3C DID-linked agent with capabilities, optional feed introduction, and get a readiness score with actionable next steps.MIT
- Records engineering-task dispatches with a full audit trail. Use when delegating parallelizable work to a subagent, capturing intent, assignee, and constraints.MIT
- Analyze a system prompt for context engineering issues and get top findings ranked by severity. Use for reviewing, improving, or debugging AI agent prompts.MIT
- Analyzes prompt template versions against labeled examples and customer feedback to generate improved versions, with optional automated testing.Apache 2.0
- Generate a formatted coding prompt for an annotation, ready to paste into Claude, Copilot, or other assistants.MIT
- Generate structured prompt templates from feature requirements or existing prompts to enable systematic testing and evaluation of AI applications.Apache 2.0
- Submit feature requests, report documentation gaps, or escalate issues to the engineering team through a direct email.ISC
- Generate an image from a text prompt. Automatically selects a checkpoint and applies defaults, then returns a prompt ID to track completion.MIT
- Update an AI agent's system prompt with new instructions. Optionally set the first message. For versioned agents, changes are saved as a draft to be published later.MIT