Leon is an open-source personal AI assistant that uses tools, memory, context, and agent-like execution to understand goals and complete tasks. People use it as a practical assistant that can run locally and work across structured workflows, while the catalogue skills support its operation.
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 leon-ai/leon --skill media-productiongit clone --depth 1 https://github.com/leon-ai/leonWrote 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/leon-ai/leon/media-production)<a href="https://agentmods.dev/skills/leon-ai/leon/media-production"><img src="https://agentmods.dev/badge/skills/leon-ai/leon/media-production/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/leon-ai/leon/media-production"><img src="https://agentmods.dev/badge/skills/leon-ai/leon/media-production.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00024 | $0.00773 |
| Opus 5.5 | $0.00010 | $0.00309 |
| Sonnet 5.5 | $0.00005 | $0.00155 |
| Haiku 4.5 | $0.00002 | $0.00077 |
Grade A, and why
media-production 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 5d 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.
What it actually says
Media Production
- Establish the desired output, duration, aspect ratio and supplied reference material. A single image or speech request usually needs only its generation tool.
- For hosted image generation, use
media_production.image.generatewith the desired prompt and output options; Core resolves tool settings, inheriting the active conversation provider by default. Ifselection_requiredis returned, ask the owner which available provider to use; never silently switch accounts. Useconfigureonly when they request a saved preference. Inspectcapabilities.defaultswhen setup needs diagnosis. ReservegenerateWithModelfor an explicitly owner-requested provider/model; never pick an older model from memory or bypass a default after an error. - Generate image assets with
media_production.image. Use existingmusic_audiospeech tools for narration and owner-supplied audio or video clips for other media assets. - For local production, use
media_production.photocraftfor layered raster edits,vectorcraftfor vector artwork,hyperframesfor social videos and motion graphics,filmcraftfor timelines, andlightcraftfor photo development. These tools run headlessly; PhotoCraft desktop control is optional. For Craft tools, discover commands and parameter schemas before edits, inspect the source, and use returned object/media IDs instead of guessing. Ordinary edits produce new artifacts while preserving originals. Keep referenced media available and use absolute asset paths. Select persistent LightCraft libraries only when the owner requests library work. - For an authored social video or motion graphic, establish a short storyboard and consistent visual direction: one main idea per scene, readable typography at phone size, deliberate spacing, a focused palette, and useful visual assets. Match pacing to the brief; avoid walls of small text, repeated generic panels and gratuitous effects. Use
hyperframes.createfor a new project and the file tool to author its HTML, CSS and scene timelines. Look for suitable catalog components before inventing complex effects. Preserve supplied source by working in a copy. Tool guidance owns the composition and animation contract. - Batch independent grounding reads and author related files together. Use
hyperframes.inspectto ground scene timing, thenhyperframes.checkwithoptions.mode=lintto catch source errors cheaply. During repairs usemode=targetedwith affectedtimesand preview neighboring frames; verify text fit, safe margins and continuity. On the final source run the default full check and requirescope=full,complete=trueandchecks.ok=true. Targeted/lint success is partial evidence. Independent checks and previews may run together once source is stable; never edit files while they are being inspected. Reuse current-run evidence and successful checks while relevant source/assets are unchanged. Render delivery quality once and review that exact video for motion, pacing and audio; use a separate draft only when useful for unresolved motion. Rerender only after changes or an observed defect. Passing checks or looking at a still frame alone cannot establish that a video is ready to share. - Preview edited artwork and representative timeline/composition positions. Use existing FFmpeg and FFprobe tools for simple composition, conversion and validation. Inspect frames and audio/duration before claiming the result is verified. Wait for background renders to complete before delivery.
- Generated and edited artifacts are attached automatically. Attach other composed/local files through
file_system.file.attachso the owner receives a player or download. Do not expose provider URLs, credentials or base64 in chat.
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.
- 5d ago Changed 04f5ccbfbde1
- 7d ago Changed · +1 lines 3b4d1247a724
- 8d ago Changed · +1 lines e49837006e13
- 16d ago First seen · 17 lines · 24 tokens per session scan A 496d00547528
media-production is a skill published in the GitHub repository leon-ai/leon (17,579 stars, last pushed 4d ago), licensed MIT. It adds 24 tokens to every session and 773 once invoked, about $0.0001 per session on Opus 5.5. 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 2026-09-30.
Other skills, from other repositories
marimo-pair
Work inside the user's live marimo notebook from the code editor: run Python in the same kernel the user does, inspect live notebook state, and commit durable notebook changes through code mode. Use whenever you create, analyze, or improve the user's marimo notebook.
fill-model-descriptions
Curate descriptions and provider-specific reasoning options and token limits in packages/llm-info/data/models.yml using models.dev, OpenRouter, and provider documentation. Use when the user asks to populate model descriptions, enrich the model catalog, or curate descriptions after running pnpm sync-models.
lepton-cli
Operate NVIDIA DGX Cloud Lepton through the globally installed lep CLI. Use when the user asks to inspect or manage Lepton workspaces, endpoints/deployments, dev pods, batch jobs, fine-tuning jobs, Ray clusters, Slurm clusters, Dynamo endpoints, storage, secrets, nodes, ingress, templates, logs, authentication, or…
flyenv-failure-boundaries
Use when implementing or reviewing FlyEnv multi-step operations, batch actions, PID candidate filtering, cleanup, supplementary actions, or changes to throw/reject and terminal result handling.
nanoresearch-experiment
Generate a Python code skeleton from an experiment blueprint.
python
The Vovk.ts Python client — generating a typed Python package from a Vovk API with vovk-python, the py / pySrc templates, TypedDict shapes, JSON + JSON Lines consumption, client-side validation, and PyPI publishing. Use whenever the user wants a typed Python SDK, ongoing Python integration, or PyPI publishing …