ffmpeg

ffmpeg is a skill for Claude Code from ai-ecoverse/slicc. It costs 79 tokens per session (1,265 once invoked), scanned A, original, Apache-2.0.

A guide to ffmpeg and ffprobe, tools for converting, editing, and inspecting audio and video files.

In plain words
What is it for?
Use it to trim or transcode video, change audio, remux containers, apply filters, create images or GIFs, and inspect media metadata.
Why use it?
It clarifies which media operations work directly and when the additional WebAssembly engine must be installed.

Skill for Claude Code

Written for Claude Code: allowed-tools in frontmatter.

Good fit Use it to trim or transcode video, change audio, remux containers, apply…

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/ai-ecoverse/slicc/ffmpeg
Install

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.

Any agent
npx skills add ai-ecoverse/slicc --skill ffmpeg
Clone the repo
git clone --depth 1 https://github.com/ai-ecoverse/slicc

Made for: Claude Code.

Wrote 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.

agentmods badge for ffmpeg

README.md
[![agentmods](https://agentmods.dev/badge/skills/ai-ecoverse/slicc/ffmpeg.svg)](https://agentmods.dev/skills/ai-ecoverse/slicc/ffmpeg)
Your own site
<a href="https://agentmods.dev/skills/ai-ecoverse/slicc/ffmpeg"><img src="https://agentmods.dev/badge/skills/ai-ecoverse/slicc/ffmpeg.svg" alt="Measured on agentmods" height="20"></a>
Per session 79 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,265 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00079 $0.01265
Opus 5 $0.00039 $0.00633
Sonnet 5 $0.00016 $0.00253
Haiku 4.5 $0.00008 $0.00127

Measured yesterday against content hash 8da7c018c238, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

ffmpeg 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 yesterday.

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.

packages/vfs-root/workspace/skills/ffmpeg/SKILL.md · 110 lines

How it starts

The opening of the file, as written. The whole thing — 110 lines — stays where its author put it; the contents beside it link to each section on GitHub.

ffmpeg / ffprobe

Both commands have two engines behind one CLI:

  • mediabunny (WebCodecs, hardware encoders, streams from disk): used automatically for one-input → one-output jobs whose every option it can express — remux, transcode to h264/hevc/vp8/vp9/av1 + aac/opus/mp3/vorbis/ flac/pcm, -ss/-t/-to trims, -vf crop,scale,fps (written in that order; transpose only on its own), -ac/-ar, -b:v/-b:a/-crf, -movflags +faststart, -metadata. Needs no install. An explicit encoder always re-encodes; -c copy never does. ffprobe answers from its container index (typed fields, no wasm boot).
  • @ffmpeg/core wasm (ipk-installed): everything else — lavfi sources, -f concat, filtergraphs (drawtext, overlay, loudnorm, …), analysis sinks (-f null), image/GIF output, codecs the browser lacks.

The choice is automatic and stderr says which engine ran. Force one with FFMPEG_ENGINE=wasm ffmpeg … (byte-identical ffmpeg behaviour) or FFMPEG_ENGINE=mediabunny ffmpeg … (fail instead of falling back, and see why). Without an explicit -c, mediabunny copies streams the container can hold rather than re-encoding them.

There is no separate ffprobe binary: on the wasm engine ffprobe runs ffmpeg -hide_banner -i <file> with no output and parses the Input #N banner. Unsupported options are rejected by name (never silently dropped).

Install (wasm engine only)

ipk add -g @ffmpeg/[email protected]

Prefer -g so scoops and the cone share one copy. Follow the pinned version printed by ffmpeg --help / ffprobe --help if it differs. There is no CDN fallback. ffmpeg -version prints a core: line saying which core is loaded.

Multi-threaded core (opt-in, single-input jobs only). On a cross-origin-isolated leader (the hosted tab on current Chrome):

ipk add -g @ffmpeg/[email protected]
FFMPEG_CORE=mt ffmpeg -i in.mp4 -c:v libx264 out.mp4

It uses several CPU cores for libx264 / libvpx and caps -threads and -filter_threads for you (pass your own to override). It deadlocks on any job with more than one input (ffmpeg's per-input demux threads), so such jobs are refused on it — run them without FFMPEG_CORE=mt. The first ffmpeg in a session decides which core stays loaded.

Read the full file on GitHub · 110 lines

Changes

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.

  1. yesterday First seen · 110 lines · 79 tokens per session scan A 8da7c018c238

Subscribe to this mod's changes

ffmpeg is a skill published in the GitHub repository ai-ecoverse/slicc (30 stars, last pushed yesterday), licensed Apache-2.0. It adds 79 tokens to every session and 1,265 once invoked, about $0.0004 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 2026-09-05.

Related

Other skills, from other repositories

audio-transcribe

Transcribe speech from audio files (mp3, m4a, wav, ogg, flac, webm) to text using the local whisper CLI — no API key. Use whenever a task hinges on the spoken content of an audio attachment.

AtomicBot-ai/atomic-agent · 58 tokens

ffmpeg

Process audio and video with the ffmpeg / ffprobe CLIs — convert, trim, extract audio, resize, change format, make GIFs, inspect media. Use for any audio/video transformation.

AtomicBot-ai/atomic-agent · 47 tokens

imagemagick

Edit and convert images with the ImageMagick magick CLI — resize, crop, convert format, compress, rotate, montage, annotate. Use for any still-image transformation.

AtomicBot-ai/atomic-agent · 41 tokens

powerpoint-presentations

Create, inspect, edit, and validate professional PowerPoint presentations (.pptx), including uploaded source decks and templates, with design-system extraction, OOXML-preserving edits, structural linting, and rendered visual QA.

aws-samples/sample-strands-agent-with-agentcore · 48 tokens

visual-design

Use this skill any time the user needs a visual output as an image or PDF — charts, diagrams, posters, infographics, abstract artwork, or any visual design. Trigger for: data visualization requests, poster/flyer creation, infographic design, abstract or artistic visuals, architecture/flow diagrams, or any request…

aws-samples/sample-strands-agent-with-agentcore · 95 tokens

lab-qc-presentation-theme

Use when creating or restyling German HTML/CSS/JS presentations for chemical production quality-control laboratory audiences: laboratory technicians, scientists, and/or quality-control teamleaders. Reproduces a modern professional green laboratory theme while regenerating content only from the current source material.

Firstp1ck/pi-coding-agent-forge · 61 tokens