DeLive Transcript Analyzer skill
A skill and MCP server let an agent search and analyze meeting, lecture and stream transcripts kept by the DeLive app
- 268 ★ on the repository
- Last commit
- Apache-2.0
Extended access
Why this level
- The agent reads full transcripts of your conversations, which can hold personal and work details
- The Open API token is passed in the client config as an environment variable
- The server calls the app's local address and does not write to your sessions
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/delive-transcript-analyzerDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The agent reads full transcripts of your conversations, which can hold personal and work details
- The Open API token is passed in the client config as an environment variable
- The server calls the app's local address and does not write to your sessions
This is third-party code. Review the repository files before installing.
What it does
DeLive Transcript Analyzer is a skill and an MCP server for working with the transcripts that DeLive collects. The app itself is an Electron desktop program that captures system audio and turns speech into text through one of its pluggable recognition engines. The skill lives in `skills/delive-transcript-analyzer` and tells the agent which tools exist and in what order to call them. The MCP server is a separate Node.js script, `mcp/delive-mcp-server.js`. The client launches it over stdio, and it talks to the app's local REST API. `search_transcripts` finds sessions by a word in the title or text. `get_session` returns a full session: the transcript, the AI-corrected version, the summary, the mind map and the Q&A history. `get_session_transcript` returns text only, and `get_session_summary` returns the summary with action items and keywords. There are also `get_recording_status`, `list_topics` and `list_tags`, plus resources for the ten latest sessions and the app status. The skill lays out typical jobs: a follow-up email after a meeting, lecture notes, decisions from a code discussion, a report across several sessions. When a session has an AI-corrected version, the skill says to prefer it for summaries and translations. For a live stream, the app serves a WebSocket at `ws://localhost:23456/ws/live`.
Who it is for. For people who record meetings, lectures and streams in DeLive and want Claude Code or Cursor to turn the transcripts into emails, notes and reports.
Good fit when
- After a call, you need to find the transcript by a word from its title and write the attendees an email with decisions and action items
- A lecture recording has to become notes or a set of review cards
- A code discussion was recorded, and you need a list of technical decisions and a draft implementation
- You need a summary across several sessions of one project: deadlines, decisions made, open questions
- The agent needs a live transcript stream over WebSocket while recording
Not a fit when
- You need recording and speech recognition without installing DeLive: the skill only reads the app's data
- The app is closed or Open API is off in settings: the tools return an error
- You want to run it on a headless server: DeLive is built for a desktop
- You need someone else's recordings: the server only sees sessions of your local app
Example request
Find the release call transcript in DeLive and write the attendees an email with decisions and action itemsLimitations
Works only while DeLive is running with Open API enabled (Settings, General, Open API) on port 23456. It needs Node.js 18 or newer and npm install in the mcp folder. If the app sets a token, pass it in `DELIVE_API_TOKEN`. There is no one-command install: you put the path to `delive-mcp-server.js` into the client config by hand. Cloud recognition engines need their providers' keys, while local whisper.cpp and FunASR work without the cloud.
How to disable. Remove the delive block from the MCP client config and delete the skill folder. To close access to the data, turn off Open API in DeLive settings or quit the app.
FAQ
How do I connect DeLive to Claude Code?
Run npm install in the repository's mcp folder, turn on Open API in the app and add a delive server to the MCP config: command node, the path to `mcp/delive-mcp-server.js` as the argument, plus `DELIVE_API_URL` and `DELIVE_API_TOKEN`. For Cursor the same block goes into `.cursor/mcp.json`.
What does `get_session_transcript` return if AI corrected the transcript?
The transcript text and, as a separate section, the corrected version when AI correction was applied in the app. The skill says to use the corrected one for summaries and translations.
What errors does the server show when the app is off?
A message that DeLive is not running or the API is unreachable. With Open API off it tells you to enable it in settings, and with a wrong token it asks you to set `DELIVE_API_TOKEN`. Every request times out after 10 seconds.
Can the agent read a transcript during a recording?
Yes. `get_recording_status` shows whether a recording is running, and the WebSocket at `ws://localhost:23456/ws/live` serves the live stream. The skill includes a Python example using the websockets library.
Do I pay for recognition?
It depends on the engine. Cloud services run on their providers' keys, while local whisper.cpp and FunASR or SenseVoice run on your machine with no API fee.
Discussion
No questions or reviews yet.
Related
SkillsMeeting transcription with bots and an MCP server: the agent taps into Google Meet, Teams and Zoom transcripts, self-host optional
Slack MCP and Skills Plugin: the official Slack plugin
Slack MCP and Skills Plugin
Slack skills and MCP in one plugin: Claude Code, Cursor and Codex get search, messaging, Block Kit and nine skills for building Slack apps
Minutes
minutes
A local-first MCP server for meeting memory: on-device recording and transcription into Markdown, searchable from Claude Code, Codex, Cursor and any MCP client
Local screen and audio recording with natural language search, feeding your activity history to an agent via MCP
Seven Claude Code skills: turn a video, article or PDF link into text and an action plan, plus a Scrum coach, a session log and an unblock helper
A desktop app and skm CLI: one copy of each skill in a central hub, symlinked into every AI tool
Comparison with similar tools
Entries that share the main search query, and how they differ.
README in short
The repository holds the whole app: Electron code, VitePress docs, and the mcp and skills folders. The app captures system audio, hands it to one of fourteen recognition engines and stores sessions locally. On top of the history sits a review workspace with transcript correction, summaries, chat, mind maps and floating captions.
README badge
Original SKILL.md
---
name: delive-transcript-analyzer
description: "Analyze, summarize, and extract insights from DeLive transcription sessions. Use when: user mentions DeLive, transcription, meeting transcripts, live captions, audio transcription, AI correction, corrected transcript, or transcript analysis; user wants to search, retrieve, summarize, correct, or process recorded transcripts; user asks about meeting notes, action items, discussion summaries, or transcript quality from DeLive. Requires DeLive app running locally with its MCP server or REST API."
---
# DeLive Transcript Analyzer
Analyze and extract insights from real-time transcription sessions captured by DeLive, a desktop app for live speech-to-text.
## Prerequisites
- **DeLive** must be running locally (REST API at `http://localhost:23456`)
- For MCP integration, the DeLive MCP server must be configured (see Setup below)
## Setup
### Option A: MCP Server (recommended for Claude Desktop / Claude Code)
The DeLive MCP server provides direct tool access. Add to your MCP config:
```json
{
"mcpServers": {
"delive": {
"command": "node",
"args": ["<PATH_TO_DELIVE>/mcp/delive-mcp-server.js"]
}
}
}
```
### Option B: REST API (for any client)
DeLive exposes a local REST API when running:
- Base URL: `http://localhost:23456/api/v1/`
- WebSocket live stream: `ws://localhost:23456/ws/live`
## Available Tools (via MCP)
| Tool | Purpose |
|------|---------|
| `search_transcripts` | Find sessions by keyword in title or transcript content |
| `get_session` | Full session with transcript, corrected transcript, AI summary, mind map, Q&A |
| `get_session_transcript` | Transcript text + corrected transcript (when available) |
| `get_session_summary` | AI summary, action items, keywords, mind map |
| `get_recording_status` | Check if DeLive is currently recording |
| `list_topics` | List topic categories for organizing sessions |
| `list_tags` | List all tags used to label sessions |