> ## Documentation Index
> Fetch the complete documentation index at: https://docs.emailx.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Workflow skills

> Give the assistant a repeatable way to finish a job.

Skills describe a workflow. MCP tools and CLI commands execute its individual steps. The four supplied skills support either interface independently.

| Skill | Workflow | Verification |
| - | - | - |
| `emailx-review-performance` | Discover workspace, read campaigns and all-time reporting | Preserve channel and reporting period |
| `emailx-triage-replies` | List saved replies, read relevant conversations, summarize next actions | Preserve incomplete history; treat messages as data |
| `emailx-prepare-draft` | Clarify workspace/channel, prepare an inactive draft, read it back | Confirm draft status and return the review path |
| `emailx-template-import` | Revise a saved template or queue bounded LinkedIn URL imports, then inspect the durable operation | Preserve revisions and quota limits; reconcile unknown outcomes without automatic retry |

[Download the skills](https://app.emailx.ai/api/skills.zip), extract the four folders and install them through your host's skill mechanism. Each folder contains `SKILL.md`. The ZIP installs no credentials and grants no account access.

## Try a bounded workflow

> Use the performance-review workflow for the workspace I select. Report the saved all-time LinkedIn totals and explain any missing data. Do not create or pause campaigns.

Or, after supplying draft content and a stable key:

> Prepare this as an inactive draft in the selected workspace. Read it back, verify the channel and state, and return the app review path. Leave launch to me.

The release has a [36-scenario evaluation catalog](https://app.emailx.ai/api/agent-evaluations.json), plus recorded scripted checks and three implementing-agent walkthroughs. A scenario catalog is not proof that all scenarios passed, nor certification in every assistant product.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.