Vellum Assistant is a personal AI assistant that remembers information about users, learns their preferences, and takes actions across connected apps. It is intended for people who want an assistant that can manage conversations, unfinished work, and proactive notifications over time. The catalogue skills, hooks, instruction, and setting configure or extend how the assistant works.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add vellum-ai/vellum-assistant --skill media-processinggit clone --depth 1 https://github.com/vellum-ai/vellum-assistantWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/vellum-ai/vellum-assistant/media-processing)<a href="https://agentmods.dev/skills/vellum-ai/vellum-assistant/media-processing"><img src="https://agentmods.dev/badge/skills/vellum-ai/vellum-assistant/media-processing.svg" alt="Measured on agentmods" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00016 | $0.03555 |
| Opus 5 | $0.00008 | $0.01777 |
| Sonnet 5 | $0.00003 | $0.00711 |
| Haiku 4.5 | $0.00002 | $0.00356 |
Grade A, and why
media-processing scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 282 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Ingest and track processing of media files (video, audio, images) through a configurable 3-phase pipeline.
End-to-End Workflow
The processing pipeline follows a sequential 3-phase flow:
- Ingest (
ingest_media) - Register a media file, detect MIME type, extract duration, deduplicate by content hash. - Preprocess (
extract_keyframes) - Detect dead time, segment the video into windows, extract downscaled keyframes, build a subject registry, and write a pipeline manifest. - Map (
analyze_keyframes) - Send each segment's frames to the configured Gemini vision model with assistant-provided extraction instructions and a JSON Schema for guaranteed structured output. Supports concurrency pooling, cost tracking, resumability, and automatic retries. - Reduce / Query (
query_media) - Send all map output to Claude for intelligent analysis and Q&A. Supports arbitrary natural language queries about video content. - Clip (
generate_clip) - Extract video clips around specific moments.
The processing pipeline service (services/processing-pipeline.ts) orchestrates phases 2-4 automatically with retries, resumability, and cancellation support.
Tools
ingest_media
Register a media file for processing. Accepts an absolute file path, validates the file exists, detects MIME type, extracts duration (for video/audio via ffprobe), and registers the asset with content-hash deduplication.
media_status
Query the processing status of a media asset. Returns the asset metadata along with per-stage progress details. Use this to monitor pipeline progress.
extract_keyframes
Preprocess a video asset: detect dead time via mpdecimate, segment the video into windows, extract downscaled keyframes at regular intervals, build a subject registry, and write a pipeline manifest.
Parameters:
asset_id(required) - ID of the media asset.interval_seconds- Interval between keyframes (default: 1s). Use 0.5s for sports/action content where frame density matters.segment_duration- Duration of each segment window (default: 15s).dead_time_threshold- Sensitivity for dead-time detection (default: 0.02).section_config- Path to a JSON file with manual section boundaries.detect_dead_time- Whether to detect and skip dead time (default: false). Dead-time detection can be too aggressive for continuous action video like sports - it may incorrectly skip live play. Enable only for content with clear idle periods (e.g., lectures, surveillance footage).short_edge- Short edge resolution for downscaled frames in pixels (default: 480).include_audio- Whether to extract and transcribe audio for each segment (default: false). When enabled, each segment's audio is transcribed using the configured STT service and stored alongside visual frames.
What ships with it
23 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- __tests__/audio-transcribe.test.ts 3.7 KB runs code
- __tests__/concurrency-pool.test.ts 2.2 KB runs code
- __tests__/cost-tracker.test.ts 2.3 KB runs code
- __tests__/extract-keyframes.test.ts 5.4 KB runs code
- __tests__/media-analysis-default-model.test.ts 5.7 KB runs code
- __tests__/preprocess-audio.test.ts 4.2 KB runs code
- __tests__/preprocess.test.ts 10 KB runs code
- services/audio-transcribe.ts 2.7 KB runs code
- services/concurrency-pool.ts 1.4 KB runs code
- services/cost-tracker.ts 2.3 KB runs code
- services/gemini-map.ts 12 KB runs code
- services/gemini-video.ts 8.3 KB runs code
- services/media-analysis-model.ts 581 B runs code
- services/preprocess.ts 19 KB runs code
- services/processing-pipeline.ts 8.1 KB runs code
- services/reduce.ts 7.6 KB runs code
- TOOLS.json 9.1 KB
- tools/analyze-keyframes.ts 5.8 KB runs code
- tools/extract-keyframes.ts 2.5 KB runs code
- tools/generate-clip.ts 8.9 KB runs code
- tools/ingest-media.ts 5.9 KB runs code
- tools/media-status.ts 2.3 KB runs code
- tools/query-media-events.ts 2.1 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago Changed aa38b7020be5
- 8d ago First seen · 282 lines · 16 tokens per session scan A 2355e860b9ae
media-processing is a skill published in the GitHub repository vellum-ai/vellum-assistant (1,201 stars, last pushed yesterday), licensed MIT. It adds 16 tokens to every session and 3,555 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
make-viral-video
Build a short news-explainer video tuned for shareability. One striking moment per video; real fetched assets; self-heal validation; pluggable TTS (Gemini-free default, OpenAI fallback).
gemini-tts
Render text to mp3 via Google Gemini Flash TTS. Free-tier eligible (1500 req/day). Use for video narration, demo voiceovers, audio notes. Parallels openai-tts; default for make-viral-video.
image-generation
Generate and edit images using Gemini Flash Image, and generate videos using Veo. Supports text-to-image, image editing, text-to-video, and image-to-video.
audio-transcribe
Transcribes audio files and voice notes to text via Gemini 2.5-flash. Integrates with Slack, Discord, and Telegram bridges so voice clips surface as readable text in tasks.
screen-record
Start or stop a screen recording of the Mac via ffmpeg. Use when the user asks to record/capture their screen to a video file (a demo, a repro, a walkthrough) — not for a single still screenshot (use macos-tools screen capture for that).
openai-tts
Render text to mp3 via OpenAI's tts-1-hd. Use for video narration, demo voiceovers, audio notes.