skills.rest
mnlt
mnltCOMMUNITY
Explore all skills by this author →
GitHub

wellread

Saves and reuses prior research notes to avoid redundant web querying via hosted MCP/JSON-RPC endpoint.

8|1|Updated Apr 21, 2026
Search Tags:#coding#json-rpc#web-research#research-management#wellread#knowledge-cache
One-click install
npx skills add https://github.com/mnlt/teleport --skill wellread
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: wellread
Source: https://github.com/mnlt/teleport/tree/main/skills/wellread
Command: npx skills add https://github.com/mnlt/teleport --skill wellread

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Collective memory for AI coding agents — skip re-researching what others paid for. Backend: MCP Streamable HTTP (JSON-RPC + session-id header), not REST.

Core Features & Use Cases

  • Memory & Karma: store and reuse research notes, track freshness and karma.
  • Pre- and post-web research workflow: use before WebSearch/WebFetch; save after deep web research to avoid duplication.
  • MCP-based backend: communicates with hosted endpoint via JSON-RPC; no local MCP server required.

Quick Start

Ask Wellread to fetch prior research before web searches and save results back to the shared cache.

Frequently Asked Questions about wellread

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I cache web research for coding agents to avoid redundant searches?▼

You can cache web research for coding agents by saving and reusing prior notes through a collective memory system, which applies before WebSearch and after deep web research to reduce duplicate querying and accelerate decisions.

Does this coding agent memory require a local MCP server to store research notes?▼

No local MCP server is required for this coding agent memory. It communicates with a hosted endpoint using MCP Streamable HTTP via JSON-RPC with a session-id header to store and retrieve research notes.

How does a knowledge cache handle research freshness for library and API questions?▼

A knowledge cache handles research freshness by maintaining specific metadata for notes, ensuring that stored information regarding library and API questions is tracked for relevance before being reused by coding agents.

When should I query a shared research cache instead of performing a web search?▼

You should query a shared research cache before performing WebSearch or WebFetch for library, API, or framework questions to check if prior research exists, preventing redundant web querying and accelerating coding decisions.

What is the best way to manage collective research memory for AI coding workflows?▼

The best way to manage collective research memory is to fetch prior research before web searches and save results back to the shared cache after deep web research, utilizing a JSON-RPC backend to track note freshness and karma.

Related Skills

Explore More Software Engineering Skills→
GabeujinGabeujin

mcp-copilot-studio-server-generator

Generate an MCP server implementation with streamable HTTP for Copilot Studio.

Community
Advanced
swarm-ai-researchswarm-ai-research

rwa-pulse

Track weekly real-world asset tokenization momentum and surface key signals.

Official
Intermediate
YrzheYrzhe
10

web-research

Search the web and retrieve current information with source attribution.

Community
Intermediate
MinhDuyDEVMinhDuyDEV

webclaw

Extract structured web content from protected or complex sites via the webclaw MCP server.

Community
Advanced