Epismo Skills

Reusable skill packages that give AI agents structured project-operation capabilities through Epismo MCP.

Why Skills

AI agents hit the same problems across teams and projects:

• Operational know-how stays trapped in chat histories.
• Prompts copy easily, but multi-step processes don't.
• Every new project restarts from scratch.

Skills solve this by packaging proven operational patterns — intake, planning, coordination, risk handling, workflow release — into portable instruction sets that any MCP-connected agent can follow.

What You Can Do

Pick a use case and run a prompt. Skills handle the rest.

Quick Start

If you're working with an AI agent, simply say:

Set up MCP and Skills from github.com/epismoai/skills

The agent will read this page and complete the steps below. Otherwise, follow the steps manually.

Complete these three steps in order. Each step depends on output from the previous one.

Step 1: Create a Secret Key

You can also create one from the UI: Settings > MCP Servers > Create Secret Key. If you do, skip to Step 2.

1-1. Request an OTP

Send your email address. A one-time PIN will be sent to that address, and the response returns an otpId.

curl -sX POST https://api.epismo.ai/v1/otp-tokens \
  -H "Content-Type: application/json" \
  -d '{"email":"you@example.com"}'
 
# => {"otpId":"<OTP_ID>"}

Save: OTP_ID from the response.

1-2. Verify the OTP and get an access token

Use the otpId from step 1-1 and the PIN delivered to your email.

curl -sX POST https://api.epismo.ai/v1/users \
  -H "Content-Type: application/json" \
  -d '{"otpId":"<OTP_ID>","pin":"<PIN_FROM_EMAIL>"}'
 
# => {"userId":"<USER_ID>","accessToken":"<ACCESS_TOKEN>"}

Save: ACCESS_TOKEN from the response. It is short-lived and used only to issue a Secret Key in the next step.

1-3. Issue a Secret Key

Use the ACCESS_TOKEN from step 1-2. The key will be scoped to a workspace.

• Personal workspace (default): omit workspaceId from the request body.
• Specific workspace: first fetch your workspace IDs, then include the target workspaceId.

# Optional: fetch workspace IDs
curl -sX GET https://api.epismo.ai/v1/workspaces \
  -H "Authorization: Bearer <ACCESS_TOKEN>"
 
# => [{"workspaceId":"<WORKSPACE_ID>","name":"..."},...]

curl -sX POST https://api.epismo.ai/v1/secret-keys \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer <ACCESS_TOKEN>" \
  -d '{"workspaceId":"<WORKSPACE_ID>"}'   # omit workspaceId to use personal workspace
 
# => {"secretKey":"<SECRET_KEY>"}

Save: SECRET_KEY from the response. It is shown only once. Export it as an environment variable before continuing:

export EPISMO_SECRET_KEY=<SECRET_KEY>

EPISMO_SECRET_KEY authenticates most Epismo API endpoints (Authorization: Bearer $EPISMO_SECRET_KEY). Only the endpoints used in this step, require the short-lived ACCESS_TOKEN.

Step 2: Add the MCP Server to Your Client

Use EPISMO_SECRET_KEY from Step 1 to configure your MCP client.

MCP endpoint:

• URL: https://mcp.epismo.ai/
• Protocol: Streamable HTTP (POST /, JSON-RPC 2.0)
• Auth header: Authorization: Bearer <SECRET_KEY>

Add the following to your client configuration:

{
  "name": "epismo-mcp",
  "url": "https://mcp.epismo.ai/",
  "headers": {
    "Authorization": "Bearer <SECRET_KEY>"
  }
}

Verify the connection by calling any MCP tool. If the server responds, the setup is complete.

Step 3: Load Skills into Your Agent

Get the skill files from this repository (clone, download, or copy — use whatever fits your environment) and load them into your agent's context.

Each skill follows this structure:

<skill-name>/
  SKILL.md          <- primary instruction file; load this into your agent's context
  references/       <- supplementary references (loaded on demand by SKILL.md)
  templates/        <- structured output templates

The setup is complete when your agent can read SKILL.md for the skills it needs.

Source of Truth

Repository: https://github.com/epismoai/skills