# Augent

## Docs

- [Agent Pipelines](https://docs.augent.app/agents/agent-pipelines.md): Augent as the intelligence layer for agentic workflows. Your MCP client orchestrates, Augent feeds, automation platforms execute.
- [Building Your First Agent](https://docs.augent.app/agents/building-your-first-agent.md): Step-by-step: capture expert knowledge, create instruction files, and run your first agentic workflow.
- [Agent Examples](https://docs.augent.app/agents/examples.md): Ready-to-adapt workflow templates for content, research, copywriting, scriptwriting, and operations.
- [Eyes & Ears](https://docs.augent.app/agents/eyes-and-ears.md): Augent turns spoken workflows into complete, replicable systems by combining audio intelligence with visual context.
- [YouTube Automation Agent](https://docs.augent.app/agents/faceless-youtube.md): Build an AI production assistant for YouTube automation. Daily briefings, script writing, multi-channel management.
- [Create Your Own Agents](https://docs.augent.app/agents/overview.md): Turn expert knowledge into agentic workflows. Capture how the best people work, then let Claude run it.
- [Architecture](https://docs.augent.app/architecture.md): How Augent processes audio, from URL to insight.
- [CLI Commands](https://docs.augent.app/cli/overview.md): Augent provides a full CLI for terminal-based workflows.
- [Changelog](https://docs.augent.app/community/changelog.md): All notable changes to Augent.
- [Discord](https://docs.augent.app/community/discord.md): Join The Heard, Augent's community Discord server.
- [Security](https://docs.augent.app/community/security.md): How to report security vulnerabilities in Augent.
- [Audio Separation](https://docs.augent.app/concepts/audio-separation.md): How Augent isolates vocals from music and background noise using Meta's Demucs v4.
- [Memory & Caching](https://docs.augent.app/concepts/memory-and-caching.md): How Augent stores transcriptions, embeddings, and diarization results so every search after the first is instant.
- [Pipeline](https://docs.augent.app/concepts/pipeline.md): How audio flows through Augent: from URL to searchable, indexed transcript.
- [Semantic Search](https://docs.augent.app/concepts/semantic-search.md): How deep_search finds content by meaning using sentence-transformer embeddings.
- [Speaker Diarization](https://docs.augent.app/concepts/speaker-diarization.md): How Augent identifies who speaks when using pyannote speaker diarization.
- [Contribution Guide](https://docs.augent.app/development/contribution-guide.md): How to contribute: code standards, PR process, and submitting your first pull request.
- [Get Help](https://docs.augent.app/development/get-help.md): Get help: join the Discord, search GitHub issues, or ask the community.
- [Local Development Setup](https://docs.augent.app/development/local-setup.md): Set up for local development. Clone the repo, install dependencies, and run tests to start contributing.
- [FAQ](https://docs.augent.app/faq.md): Common questions about using and troubleshooting Augent.
- [Claude Code Setup](https://docs.augent.app/getting-started/claude-code-setup.md): Connect Augent to Claude Code so Claude can download, transcribe, and search audio natively.
- [Codex Setup](https://docs.augent.app/getting-started/codex-setup.md): Connect Augent to OpenAI Codex so it can download, transcribe, and search audio natively.
- [Deep Search](https://docs.augent.app/getting-started/deep-search.md): Search audio by meaning. Find what was said even when you don't know the exact words.
- [Download Audio](https://docs.augent.app/getting-started/download-audio.md): Audio extraction from any URL, engineered for speed and scale.
- [Installation](https://docs.augent.app/getting-started/installation.md): What the installer does, how to install manually, and how to uninstall.
- [MCP Client Setup](https://docs.augent.app/getting-started/mcp-client-setup.md): Connect Augent to any MCP-compatible client — Cursor, Windsurf, VS Code, Cline, and more.
- [Memory](https://docs.augent.app/getting-started/memory.md): How Augent remembers everything your agents have ever heard.
- [OpenClaw Setup](https://docs.augent.app/getting-started/openclaw-setup.md): Let your lobster agents master any content in seconds.
- [Quick Start](https://docs.augent.app/getting-started/quick-start.md): Install Augent and start searching audio in under 2 minutes.
- [Take Notes](https://docs.augent.app/getting-started/take-notes.md): One URL in, formatted notes out. The end of manual note-taking.
- [Clips & Highlights](https://docs.augent.app/guides/clips-and-highlights.md): Find the best moments in any audio and export them as video clips.
- [Configuration](https://docs.augent.app/guides/configuration.md): Set default values for model size, output directories, and more. Disable tools you don't need.
- [Obsidian Setup](https://docs.augent.app/guides/obsidian-setup.md): Make every .txt and .md file on your Mac open directly in Obsidian.
- [Source Separation](https://docs.augent.app/guides/source-separation.md): Isolate vocals from music, background noise, and other sounds. Clean audio in, clean transcription out.
- [Workflows & Outputs](https://docs.augent.app/guides/workflows.md): Route Augent output to notes apps, automate pipelines, and export in any format.
- [Troubleshooting](https://docs.augent.app/help/troubleshooting.md): Common issues and how to fix them.
- [ ](https://docs.augent.app/index.md)
- [The Memory Architecture](https://docs.augent.app/obsidian/memory-architecture.md): How augent's three-layer memory system gives Claude persistent, searchable, compounding knowledge.
- [Obsidian Integration](https://docs.augent.app/obsidian/overview.md): Turn your knowledge into a visual graph
- [Tips](https://docs.augent.app/obsidian/tips.md): Get the most out of your Augent + Obsidian workflow
- [Vault Setup](https://docs.augent.app/obsidian/vault-setup.md): Use Augent with your existing Obsidian vault
- [Why Augent + Obsidian](https://docs.augent.app/obsidian/why-augent-obsidian.md): Your AI stops forgetting. Your notes start compounding. Here's why this stack matters.
- [Workflows](https://docs.augent.app/obsidian/workflows.md): Daily routines and practical workflows for Obsidian + Claude Code + augent.
- [Python API](https://docs.augent.app/python-api/overview.md): Use Augent directly in Python scripts and applications.
- [batch_search](https://docs.augent.app/tools/batch-search.md): Search multiple audio files for keywords in parallel.
- [chapters](https://docs.augent.app/tools/chapters.md): Auto-detect topic chapters in audio with timestamps.
- [clear_memory](https://docs.augent.app/tools/clear-memory.md): Wipe all stored transcriptions, embeddings, diarization data, and markdown files.
- [clip_export](https://docs.augent.app/tools/clip-export.md): Export a video clip from any URL for a specific time range. Downloads only the requested segment, not the full video.
- [download_audio](https://docs.augent.app/tools/download-audio.md): Extract audio from any URL at maximum speed.
- [highlights](https://docs.augent.app/tools/highlights.md): Export MP4 clips of specific moments. Auto mode picks top moments. Focused mode finds moments matching a specific topic, person, or concept.
- [identify_speakers](https://docs.augent.app/tools/identify-speakers.md): Identify who speaks when in audio. Uses pyannote-audio for state-of-the-art speaker diarization.
- [list_files](https://docs.augent.app/tools/list-files.md): List media files in a directory. Finds all common audio and video formats by default.
- [list_memories](https://docs.augent.app/tools/list-memories.md): List all stored transcriptions with their titles, durations, dates, and file paths.
- [memory_stats](https://docs.augent.app/tools/memory-stats.md): See how many files Augent remembers, total hours stored, and disk usage.
- [take_notes](https://docs.augent.app/tools/notes.md): One URL in, formatted notes out.
- [rebuild_graph](https://docs.augent.app/tools/rebuild-graph.md): Rebuild Obsidian graph view data for all transcriptions
- [search_audio](https://docs.augent.app/tools/search-audio.md): Search audio files for keywords and return timestamped matches with context snippets.
- [search_memory](https://docs.augent.app/tools/search-memory.md): Search across ALL stored transcriptions by keyword or meaning. No audio_path needed, queries everything in memory.
- [search_proximity](https://docs.augent.app/tools/search-proximity.md): Find where one keyword appears near another keyword in audio.
- [deep_search](https://docs.augent.app/tools/semantic-search.md): Search audio by meaning, not just keywords. Uses sentence-transformer embeddings to find segments that match a natural language query.
- [separate_audio](https://docs.augent.app/tools/separate-audio.md): Separate audio into stems (vocals, drums, bass, other) using Meta's Demucs v4. Isolates vocals from music, background noise, and other sounds for clean transcription.
- [spaces](https://docs.augent.app/tools/spaces.md): Download or live-record X/Twitter Spaces audio. Auto-detects live vs ended Spaces. One tool, three modes — start a download, check status, or stop a live recording.
- [tag](https://docs.augent.app/tools/tag.md): Add, remove, or list tags on transcriptions. Tags are broad topic categories for organizing and filtering memories.
- [text_to_speech](https://docs.augent.app/tools/text-to-speech.md): Convert text to natural speech audio using Kokoro TTS. 54 voices across 9 languages, fully local.
- [transcribe_audio](https://docs.augent.app/tools/transcribe-audio.md): Transcribe an audio file and return the full text with timestamps. Results are stored in memory automatically.
- [visual](https://docs.augent.app/tools/visual.md): Extract visual context from video at moments where audio alone isn't enough. Query mode finds specific moments. Auto mode detects them autonomously. Assist mode flags visual gaps for manual screenshots.
- [Web UI](https://docs.augent.app/web-ui/overview.md): Browser-based transcription and search. Runs 100% locally, no data leaves your machine.

## OpenAPI Specs

- [openapi](https://docs.augent.app/api-reference/openapi.json)