Most agents are powerful in demos, limited in practice. We give agents the Skills to handle real work.
Everything your agent needs to run reliably and perform at its best
01
Find best practices
Search community workflows and quickly bring proven ways of working into your projects.
02
Capture your know-how
Turn your practical expertise into reusable workflows for yourself, your team, or the community.
03
Operate as projects
Connect imported workflows to projects and execute, track, and manage them as ongoing tasks.
Quick start
Ask your agent to set it up and you're ready to go.
Set up Epismo access and Skills from github.com/epismoai/skills. I can use MCP or CLI.
MCP server
Connect any MCP-compatible tool to your Epismo workspace over a single endpoint. Your agent gets direct access to shared context, tasks, and so on.
Endpoint
Add this endpoint in your MCP-compatible tool to connect to Epismo.
Available tools (epismo_{resource}_{action})
- epismo_pack_{search|get|create|update|delete|like}
- epismo_track_{search|get|create|update|delete|apply}
CLI
Interact with your workspace directly from the terminal — the natural home for agent-driven automation. Search packs and manage tracks without opening a browser.
Get started
Install the CLI globally to start using Epismo from your terminal.
npm install -g epismo
Available commands (epismo {resource} {action})
- epismo pack {search|get|create|update|delete|like}
- epismo track {search|get|create|update|delete|apply}
Files
LICENSE
README.md
skills/
context-pack/
references/
search.md
visibility.md
SKILL.md
templates/
content.md
epismo-basics/
references/
credit-purchase.md
pack-alias.md
SKILL.md
project-tracking/
references/
ai-delegation.md
runbook.md
search.md
SKILL.md
templates/
operations.md
workflow-pack/
references/
quality.md
release.md
search.md
visibility.md
SKILL.md
templates/
patterns.md
Epismo Skills
Reusable skill packages that give AI agents structured capabilities through Epismo MCP or CLI.
Why Skills
AI agents hit the same problems across teams and tools:
- Operational know-how stays trapped in chat histories.
- Multi-step processes don't transfer when you switch tools.
- Every new project restarts from scratch.
Skills solve this by packaging proven operational patterns into portable instruction sets that any agent with Epismo access can follow.
Skills
| Skill | What it does |
|---|---|
| Epismo Basics | Platform fundamentals: auth, CLI/MCP conventions, scope, share URL resolution, error handling |
| Project Tracking | Create and update tasks and goals; plan multi-step work; unblock stalled queues |
| Workflow Pack | Discover, adapt, and release reusable workflows |
| Context Pack | Save session context, hand off tasks, load saved context from any tool |
Epismo Basics is a shared foundation — load it alongside any other skill.
Use Cases
| Goal | Skills to load |
|---|---|
| Resume work after switching tools | Context Pack |
| Hand off a task to a teammate | Context Pack |
| Add tasks, update status, plan a sprint | Epismo Basics + Project Tracking |
| Find and reuse a community workflow | Epismo Basics + Workflow Pack |
| Capture a proven process and publish it | Epismo Basics + Workflow Pack |
| Delegate work to an AI agent with clear criteria | Epismo Basics + Project Tracking |
| Publish a best-practice guide for the community | Context Pack |
| Full project operations | Epismo Basics + Project Tracking + Workflow Pack |
Quick Start
Tell your agent:
Set up Epismo access and load the Skills from https://raw.githubusercontent.com/epismoai/skills/main/README.md.
The agent will read this page and complete the steps. Or follow manually:
1. Connect
CLI and MCP connect to the same Epismo service. Use CLI if available; MCP otherwise.
CLI (preferred):
npm install -g epismo
epismo login --email you@example.com
epismo whoami
MCP: add https://mcp.epismo.ai as an MCP server in your client. Authentication is handled automatically via OAuth.
2. Select a workspace
epismo workspace list
epismo workspace use --workspace-id <workspace-id>
3. Load skills
Get the skill files from this repository (clone, download, or copy) and load the relevant SKILL.md files into your agent's context. Each skill follows this structure:
<skill-name>/
SKILL.md ← load this
references/ ← loaded on demand
templates/ ← structured output templates
Proven AI use cases, ready to use.