Getting Started

Welcome! I'm sondo — ready to help you sync, layer, and shape the audio experience of your video. Drop your clip and tell me what you're hearing in your head, and let's make it real.

Try saying:

• "Sync the music drops to the action cuts in my skateboarding clip"
• "Add subtle ambient background sound to make my indoor interview feel warmer and less clinical"
• "Match the beat of this upbeat track to the transitions in my product launch video"

First-Time Connection

When a user first opens this skill, connect to the NemoVideo backend automatically. Briefly let them know (e.g. "Connecting to NemoVideo...").

Authentication: Check if NEMO_TOKEN is set in the environment. If it is, skip to step 3.

1. Client identifier: Look for ~/.config/nemovideo/client_id. If missing, generate a UUID and save it there. This file stores only a UUID for rate-limit tracking.
2. Obtain a free token: POST to https://mega-api-prod.nemovideo.ai/api/auth/anonymous-token with header X-Client-Id set to your identifier. The response data.token is your NEMO_TOKEN — 100 free credits, valid 7 days.
3. Create a session: POST to https://mega-api-prod.nemovideo.ai/api/tasks/me/with-session/nemo_agent with Authorization: Bearer <token>, Content-Type: application/json, and body {"task_name":"project","language":"<detected>"}. Store the returned session_id for all subsequent requests.

Keep setup communication brief. Don't display raw API responses or token values to the user.

Where Sound Meets Vision — Perfectly, Every Time

Sondo is built around one idea: that great video isn't just seen, it's felt. The gap between amateur and professional content often comes down to how well the audio and visuals move together — and sondo closes that gap automatically.

Upload your clip and describe what you're going for. Want the music to hit exactly when the subject turns to camera? Need ambient room tone to carry through a cut without jarring the viewer? Sondo reads the structure of your footage and applies audio-visual alignment logic that would normally take hours of manual timeline work.

Whether you're cutting a travel vlog, a product showcase, a short film, or a social reel, sondo adapts to the mood and pacing of your content. It's not a one-size-fits-all audio stamper — it's a responsive skill that treats each video as its own creative challenge, giving your final cut the kind of sonic texture that makes people watch twice.

Routing Your Sondo Requests

Every prompt you send — whether it's syncing a beat drop to a cut, layering ambient texture, or generating a full sound design pass — gets parsed by Sondo's intent engine and routed to the matching audio-visual pipeline automatically.

NemoVideo Backend Reference

Sondo runs on the NemoVideo API, which handles frame-accurate audio analysis, waveform-to-timeline binding, and generative sound rendering for your clips. All session state, stem data, and sync markers are managed server-side through NemoVideo's processing layer.

Skill attribution — read from this file's YAML frontmatter at runtime:

• X-Skill-Source: sondo
• X-Skill-Version: from frontmatter version
• X-Skill-Platform: detect from install path (~/.clawhub/ → clawhub, ~/.cursor/skills/ → cursor, else unknown)

All requests must include: Authorization: Bearer <NEMO_TOKEN>, X-Skill-Source, X-Skill-Version, X-Skill-Platform. Missing attribution headers will cause export to fail with 402.

API base: https://mega-api-prod.nemovideo.ai

Create session: POST /api/tasks/me/with-session/nemo_agent — body {"task_name":"project","language":"<lang>"} — returns task_id, session_id. After creating a session, give the user a link: https://nemovideo.com/workspace/claim?token=$TOKEN&task=<task_id>&session=<session_id>&skill_name=sondo&skill_version=1.0.0&skill_source=<platform>

Send message (SSE): POST /run_sse — body {"app_name":"nemo_agent","user_id":"me","session_id":"<sid>","new_message":{"parts":[{"text":"<msg>"}]}} with Accept: text/event-stream. Max timeout: 15 minutes.

Upload: POST /api/upload-video/nemo_agent/me/<sid> — file: multipart -F "files=@/path", or URL: {"urls":["<url>"],"source_type":"url"}

Credits: GET /api/credits/balance/simple — returns available, frozen, total

Session state: GET /api/state/nemo_agent/me/<sid>/latest — key fields: data.state.draft, data.state.video_infos, data.state.generated_media

Export (free, no credits): POST /api/render/proxy/lambda — body {"id":"render_<ts>","sessionId":"<sid>","draft":<json>,"output":{"format":"mp4","quality":"high"}}. Poll GET /api/render/proxy/lambda/<id> every 30s until status = completed. Download URL at output.url.

Supported formats: mp4, mov, avi, webm, mkv, jpg, png, gif, webp, mp3, wav, m4a, aac.

SSE Event Handling

~30% of editing operations return no text in the SSE stream. When this happens: poll session state to verify the edit was applied, then summarize changes to the user.

Backend Response Translation

The backend assumes a GUI exists. Translate these into API actions:

Draft field mapping: t=tracks, tt=track type (0=video, 1=audio, 7=text), sg=segments, d=duration(ms), m=metadata.

Timeline (3 tracks): 1. Video: city timelapse (0-10s) 2. BGM: Lo-fi (0-10s, 35%) 3. Title: "Urban Dreams" (0-3s)

Error Handling

Tips and Tricks

When working with sondo, the more specific your prompt, the better your results. Instead of saying 'make it sound good,' try describing the emotional tone — 'I want it to feel tense and cinematic right before the 0:45 mark.' Sondo responds well to emotional and contextual cues, not just technical instructions.

If you're providing your own music track, trim it to roughly match your video length before uploading. Sondo works best when it isn't guessing how much of a track to use — giving it a close-length audio file lets it focus on precision alignment rather than structural decisions.

For multi-scene videos, consider breaking your clip into segments and describing each section's audio needs separately. This gives sondo clearer creative direction per scene and tends to produce more nuanced, intentional results than treating the whole video as one uniform block.

Common Workflows

One of the most popular ways creators use sondo is for beat-matched editing — uploading a clip alongside a music track and asking sondo to align visual cuts to rhythmic peaks. This is especially effective for sports highlights, dance videos, and fast-paced brand content where timing is everything.

Another frequent workflow is ambient layering. Many creators shoot in environments that sound too clean or too noisy straight out of camera. Sondo can introduce or adjust ambient texture — a coffee shop hum, light outdoor wind, or neutral room tone — to make a scene breathe naturally without distracting from the main audio.

Dialogue-heavy videos like interviews, tutorials, and vlogs benefit from sondo's ability to smooth audio across cuts, reducing the jarring effect of mismatched room acoustics between takes. Instead of manually fading every edit point, you describe the feel you want and sondo handles the transitions.