name: redreplier title: RedReplier repo: https://github.com/RedReplier/claude-plugin sha: e475c10b0b9b9ed53c91fb756f6c334d829460b1 version: "1.4.2" description: "Find leads in public conversations through RedReplier's hosted MCP server (Streamable HTTP at https://mcp.redreplier.com/mcp, OAuth sign-in, no keys in config): keyword mentions on Reddit, Hacker News, X, Bluesky and Facebook, each scored 0-100 for relevance, with filters by score, keyword, source and date, the reason behind a score, approve and reject triage, monitored websites and keywords, and alert settings. Bundles the redreplier workflow skill. Agent Plugins v1 package (plugin.json + mcp.json + skills/), configuration only. Disclosure — every tool call goes to RedReplier's hosted server (mcp.redreplier.com) and acts on the user's RedReplier account with their own workspace role; the tool set is served by that server at runtime and is not covered by the pinned sha. Keywords count against the user's RedReplier plan. Sign-in runs the pinned mcp-remote@0.1.49 through npx (Node.js 20.18.1+ and npm required), which opens a browser and a local callback port on first use and stores OAuth tokens under ~/.mcp-auth. The package marks the server untrusted, so Hermes asks before every tool the server does not mark read-only and refuses them in unattended runs." maintainer: RedReplier tier: community category: tools docs_url: https://github.com/RedReplier/claude-plugin#install-in-hermes-agent requires_hermes: ">=0.21.6" readme: true platforms: [] capabilities: provides_tools: [] provides_hooks: [] provides_middleware: [] requires_env: []