abhichandra21/promptheus
skills/promptheus/SKILL.md
AI-powered prompt refinement and improvement using the Promptheus MCP server. Use this skill whenever the user asks to refine, improve, rewrite, or optimize a prompt. Trigger phrases include "refine this prompt", "improve this prompt", "make this prompt better", "optimize this prompt", "rewrite this prompt", or any request that involves taking an existing prompt/instruction and making it more effective. Also triggers when the user provides rough ideas, requirements, or specifications and asks for a well-crafted prompt to be generated from them.
tools
Updated Mar 29, 2026
$ install --global
skillsauthnpx skillsauth add abhichandra21/dotclaude promptheusInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Apr 1, 2026, 12:02 AM57.1s1 file scanned
SKILL.md
- name:
- promptheus
- description:
- AI-powered prompt refinement and improvement using the Promptheus MCP server. Use this skill whenever the user asks to refine, improve, rewrite, or optimize a prompt. Trigger phrases include "refine this prompt", "improve this prompt", "make this prompt better", "optimize this prompt", "rewrite this prompt", or any request that involves taking an existing prompt/instruction and making it more effective. Also triggers when the user provides rough ideas, requirements, or specifications and asks for a well-crafted prompt to be generated from them.
Promptheus Prompt Refinement
Route all prompt refinement requests through the Promptheus MCP tools. Never attempt to refine prompts using your own capabilities -- always delegate to Promptheus.
Workflow
Refine a prompt
- Call
mcp__promptheus__refine_promptwith the user's prompt text. - If the response
typeisclarification_needed:- Use
AskUserQuestionwith the questions fromquestions_for_ask_user_question. - Map user answers to question IDs (
q0,q1,q2, ...). - Call
mcp__promptheus__refine_promptagain with the original prompt and theanswersdict.
- Use
- If the response
typeisrefined:- Present the refined prompt to the user.
- If the user originally asked to "execute" or "run" the prompt, use the refined prompt to generate the requested content.
Tweak an existing prompt
When the user wants a targeted adjustment to an already-refined prompt (e.g., "make it shorter", "more technical", "add constraints"):
- Call
mcp__promptheus__tweak_promptwith the current prompt and the requested modification. - Present the tweaked result to the user.
Rules
- Never refine or rewrite prompts yourself. Always use Promptheus tools.
- If Promptheus returns an error, report it to the user and offer to retry or troubleshoot with
mcp__promptheus__validate_environment.
Related Skills
abhichandra21/writing-assistant
documentation
VerifiedTrustedCommunity
Rephrase, correct, and expand writing with context-aware improvements. Use when the user needs help with (1) Rephrasing text for clarity, tone, or style improvements, (2) Correcting grammatical errors, typos, or spelling mistakes, (3) Expanding rough ideas, bullet points, or fragments into full prose. Trigger phrases include "rephrase this", "fix this text", "help me write", "expand this idea", or simply providing text that needs improvement.
SKILL.mdUpdated Mar 29, 2026
abhichandra21/review-board
tools
VerifiedTrustedCommunity
Multi-LLM review board. Sends any document (architecture docs, design docs, RFCs, API specs, runbooks) to external LLM CLIs (Codex, Gemini, GitHub Copilot) for independent review, synthesizes feedback, then optionally drives rebuttal rounds and consensus building. Multi-round, each round is opt-in. Trigger phrases include "review this doc", "get external review", "send to chatgpt and gemini", "review board", or any request to have a document reviewed by other LLMs.
SKILL.mdUpdated Mar 29, 2026
abhichandra21/review-board-copilot
tools
VerifiedTrustedCommunity
Multi-model review board powered by GitHub Copilot CLI. Sends any document to multiple AI models (GPT, Claude, Gemini) via a single Copilot CLI for independent review, synthesizes feedback, then optionally drives rebuttal rounds and consensus building. Uses Copilot subscription tokens instead of separate API keys. Trigger phrases include "copilot review", "review board copilot", "multi-model review", "debate this", "copilot debate", or any request for Copilot-powered document review or focused design debate.
SKILL.mdUpdated Mar 29, 2026
abhichandra21/code-review-copilot
tools
VerifiedTrustedCommunity
Multi-model code review board powered by GitHub Copilot CLI. Reviews PRs, branches, commits, or uncommitted changes by sending review targets to multiple AI models (GPT, Claude, Gemini) via Copilot CLI. Models fetch diffs and browse the codebase independently, then Claude synthesizes findings through optional rebuttal and consensus rounds. Uses Copilot subscription tokens instead of separate API keys. Trigger phrases include "code review copilot", "copilot code review", "review this PR", "review my changes", "review this branch", "copilot debate code", or any request for Copilot-powered code review or code-focused design debate.
SKILL.mdUpdated Mar 29, 2026