Copilot Instructions for Solana Metrics MCPServer
This project is a Model Context Protocol (MCP) server specifically designed for analyzing Solana blockchain metrics from InfluxDB and generating Grafana dashboards.
## Project
217mo agoReviewMountCreate a free account to mount
GitHub Copilot.
- For coding, review, and refactor tasks, follow `.agents/skills/karpathy-guidelines/SKILL.md`.
- This project is an MCPserver for browser automation using Playwright and `browser-use`.
- Runtime code lives in `src`; `main.py
base is pre-scaffolded — run `/clone-website [ ...]`.
**Cursor-first:** Prefer Cursor’s built-in browser MCP (`cursor-ide-browser`) when available. Also works with Claude Code, Codex, Copilot, Windsurf, and other
everspring_mcp\mcp\server.py`) is SDK-native (`mcp.server.Server`) and publishes one MCP tool: `search_spring_docs`.
- The server uses late initialization and startup preheating in lifespan hooks (`prefetch_model()` + `ensure
mcp.json` (worktree sessions) → use the CLI twin `task_graph_cli.py --json`; Hermes operators add the proposed `mcp_servers:` entry to `~/.hermes/config.yaml`.
## Non-negotiable invariants
Each rule's full text is `.ai-badger
mcp.json` (worktree sessions) → use the CLI twin `task_graph_cli.py --json`; Hermes operators add the proposed `mcp_servers:` entry to `~/.hermes/config.yaml`.
## Non-negotiable invariants
Each rule's full text is `.ai-badger
reviewer)
This repository is reviewed by **GitHub Copilot Coding Agent** with
the ARC-1 MCPserver attached. When you (Copilot) are invoked on a
pull request, follow these instructions.
## Your
Smoke test against a running VM
python tests/smoke_test.py
# Run the demo
bunkervm demo
# Start MCPserver (stdio for Claude Desktop, sse for web)
python -m bunkervm # stdio
python -m bunkervm
while user-facing help and explanations should be in Korean
## Architecture & Transport Selection
This MCPserver supports dual transport modes with flexible configuration via CLI args and environment variables:
- **stdio
Copilot Instructions for AI Agents
## Project Overview
- **Redfish MCPServer**: Modular Python server for controlling Redfish-enabled hardware via Model Context Protocol (MCP).
- **Main entry**: `src/redfish_mcp_server/main.py` orchestrates tool registration
115mo agoReviewMountCreate a free account to mount
data/` folder.
- Email (Gmail), calendar, and LinkedIn publishing/fetching are provided by the
`connected-workspace` MCPserver (`.agents/mcp/connected-workspace.json`). Prefer those MCP
tools over writing new scripts for those capabilities.
- Shared artifact
CostMax
CostMax is a local-first efficiency layer for coding agents: a Go CLI +
MCPserver that reduces large tool output for Codex while preserving full
evidence locally.
## Ground rules
Memory Bank extension. Edit freely to customize.
This project uses the Memory Bank MCPserver to persist context across AI sessions.
You have access to Memory Bank MCP tools
sequentially without interruption, keep the Jira board updated in real-time via the Atlassian MCPserver, and document the final delivery.
## General Principles
- **Complete Autonomy**: Once the sprint is launched
Contract.
- Contract code: Rust/Soroban in `contracts/paygate-registry`.
- Frontend: React + TypeScript + Vite in `apps/web`.
- API and MCPserver: TypeScript in `apps/api`.
README content must stay project-focused and must not include
thin adapter — it exposes both the `runmark` CLI and the `runmark mcp` stdio MCPserver
- keep tracked request intent in `runmark/` and runtime state in `runmark/artifacts/`
- do not add secret
instructions below are written for any agent. The phone is driven
> through an MCPserver, so any client that speaks MCP can use it: the
> tools appear here under