Aishelf Mcp Server
Let your coding agent pull the right AIShelf workflows, rules, skills, and prompts for each task instead of guessing from static context.
Overview
AIShelf MCP server is a Build-phase MCP integration that lets an AI model choose AIShelf workflows, rules, skills, and prompts for a task.
What is this MCP server?
- Exposes AIShelf workflows, rules, skills, and prompts via MCP stdio transport
- npm package @aishelf/mcp-server at registry version 1.0.3
- Requires YOUR_API_KEY secret for authenticated AIShelf access
- Lets the model select task-appropriate assets rather than one fixed system prompt
- Pairs with the AIShelf IDE extension repository under mcp-server subfolder
- npm package @aishelf/mcp-server version 1.0.3
- required secret env YOUR_API_KEY
What problem does it solve?
Agents burn context repeating the wrong prompts because they cannot browse your organized AIShelf library during a task.
Who is it for?
Solo builders who centralize prompts and skills in AIShelf and want MCP-driven pick-and-apply during agent sessions.
Skip if: Teams with no AIShelf account or builders who rely solely on repo-local SKILL.md files without an external catalog.
What do I get? / Deliverables
After install, the agent can request AIShelf-backed workflows, rules, skills, and prompts matched to the current work.
- MCP tools that query AIShelf workflows, rules, skills, and prompts
- Task-aligned asset selection inside agent sessions
- Authenticated stdio connection at server version 1.0.3
Recommended MCP Servers
Journey fit
Canonical shelf is Build because the server wires external prompt and skill libraries into the agent at implementation time. Agent-tooling is where MCP bridges connect assistants to curated capability catalogs like AIShelf.
How it compares
AIShelf catalog bridge via MCP, not a standalone agent skill file in your repository.
Common Questions / FAQ
Who is aishelf-mcp-server for?
It is for developers using AIShelf who want their MCP agent to select workflows, rules, skills, and prompts per task.
When should I use aishelf-mcp-server?
Use it during Build when you want dynamic, task-specific assets from AIShelf instead of a single static system prompt.
How do I add aishelf-mcp-server to my agent?
Install @aishelf/mcp-server from npm, configure stdio MCP in your agent, and set the required YOUR_API_KEY environment variable.