# Essence Ai Essence MCP server

Source: https://glimind.com/server/mcp-registry%2Forg.essence-ai%2Fessence
Registry id: `mcp-registry/org.essence-ai/essence`
Status: live (checked 2026-11-22T14:59:57.274Z)

## Measurements (30 days)

- Uptime: 100%
- Checks: 112
- Outages (90 days): 0

## Install

```json
{
  "mcpServers": {
    "essence-ai-essence-mcp-server": {
      "type": "http",
      "url": "https://mcp.essence-ai.org/mcp"
    }
  }
}
```

Claude Code:

```sh
claude mcp add --transport http essence-ai-essence-mcp-server https://mcp.essence-ai.org/mcp
```

## Tools (38)

- `account_status`: Subscription and connection status for the current account; excludes secrets.
- `add_sources`: Add public channels or import folders by link; uses the bot's shared limits.
- `check_source_availability`: Check a connected public channel's current protection and availability through Telegram. Updates its observed status but never resumes processing or delivery. I
- `configure_digest`: Configure a digest. Set consent=true only after explicit consent to collection and AI processing. External delivery from a private source requires separate priv
- `configure_hashtag`: Create/delete a hashtag or link/unlink it to a source in the current account.
- `configure_publication`: Manage an automatic publication destination and its queue. latest/activate/resume may send posts. settings: summary_language, summary_enabled, photo, video, aud
- `connect_publication_channel`: Connect an automatic publication destination only after explicit user confirmation (confirmed=true). request sends a channel/group chooser to the token owner's 
- `disable_digest_schedule`: Return to on-demand mode after the user's confirmation. Cancels future scheduled runs and unsent scheduled parts; collection and manual requests remain availabl
- `get_ai_settings`: Custom instructions and selected model; keys are never returned. Custom instructions apply only when connection.ready is true; otherwise standard Essence summar
- `get_digest_run`: Request status and, with include_result, the ready digest after rechecking access.
- `get_digest_schedule`: Read on-demand/daily/weekly mode, local time, IANA timezone, recipients, next run and its period, and pause status for a digest in this account.
- `get_free_settings`: Read ongoing Free plan availability and selected sources and destination.
- `get_publication_replacements`: Read original-post replacement rules for one source and publication destination. Rules run in list order: {kind: text|url, find: str, replace: str, enabled: boo
- `get_publication_signatures`: Read a destination's signature text and top/bottom position, plus all source and digest overrides. Override modes: inherit (use destination), custom, off (no si
- `get_settings`: Global settings or source overrides and their effective values.
- `list_ai_models`: Available saved models for the account's own AI connection.
- `list_digests`: Group and channel digests, delivery destinations and shared quota.
- `list_hashtags`: Hashtags for the current account.
- `list_publication_channels`: Automatic publication destinations, default settings and per-source overrides. In source_overrides, null media values and source_filter inherit the destination 
- `list_publications`: Publication status for a destination in the current account; excludes original content.
- `list_sources`: Connected sources, delivery pause status, availability and hashtags.
- `open_account_action`: Open a protected action. Secret links are sent to the account owner in Telegram.
- `preview_ai`: Preview processing of the latest post. The AI request is billed to your API key. Requires user confirmation; the result is also delivered to Telegram.
- `preview_digest_schedule`: First step: show the first scheduled run, its period and delivery destinations without enabling it. Ask the user for an explicit IANA timezone; never infer it f
- `preview_publication_replacements`: Check draft rules using the replacement contract from get_publication_replacements. Pass plain text or a saved processed_message_id; omit both to use the latest
- `read_feed`: Ready summaries and personal AI results. include_original adds original text. Private original text and search require the additional posuti:read_private_origin
- `refresh_ai_models`: Check the account's own AI connection and refresh the provider's model list.
- `remove_source`: Remove a source from the current account after the user's confirmation.
- `request_digest`: Request a digest for a timezone-aware ISO 8601 period; consumes the bot's shared quota. The existing queue prepares the result. Use get_digest_run for status an
- `resume_source`: Explicitly re-enable an unavailable public source after the user's confirmation. Rechecks Telegram access, current plan access and a free source slot. Resumes o
- `select_ai_model`: Select an available model for the account's own AI connection.
- `select_free_settings`: Select up to three public sources from the current account and one Free destination. Requires user confirmation; other settings are preserved while paused.
- `set_digest_schedule`: Second step: enable or resume exactly the schedule returned by preview_digest_schedule. Set confirmed and timezone_confirmed only after the user confirms the pr
- `set_publication_replacements`: Replace the entire ordered rule list only after user confirmation (confirmed=true). Use the rule contract and revision from get_publication_replacements; [] rem
- `set_publication_signature`: Set or remove a destination signature or one source/digest override after user confirmation. Omit source_id and digest_id for the destination default; otherwise
- `set_source_delivery`: Pause or resume personal delivery for a source.
- `update_ai_settings`: Update instructions {mode: inherit|custom|standard, prompt: str}. Never pass an API key here.
- `update_settings`: Update summary_enabled, media {photo,video,audio,document,video_note,poll}, source_filter; global settings also support language, summary_style, notifications_e

Measured by Glimind (https://glimind.com): MCP handshake + tool list only, no tool is ever called. CC BY 4.0.
