Guide · Agent skill
How to use the Photo to Anime Restyler agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The Photo to Anime Restyler is an agent skill: a small add-on that gives an AI coding agent such as Claude Code, Codex or OpenClaw a new capability. With it installed, you can ask the agent to turn your photos into anime, manga, comic, cartoon, watercolor, clay or 3D character art. The rendering runs on the paid, hosted Beatra service, not on your computer, so this guide explains what gets sent, what it costs and how to install and approve each step. It is written for readers who have never used agent skills before.
Photo to Anime Restylersickn33/agentic-awesome-skills · MIT Download the skill Quick start
- 1Check the requirements. Confirm you have Git, Python 3.10 or newer, a Beatra account with credits and a browser for sign-in.
- 2Install the skill. Download the ZIP from the directory page and unzip it into ~/.claude/skills/ or a project's .claude/skills/ folder, or upload it under Settings, Capabilities, Skills in the Claude apps.
- 3Review the package. Read every file in the package, including the scripts, before you activate it, and approve the copy as a separate decision.
- 4Run the free connection check. Run the verify script first. It is non-billable and confirms the connection works before any render.
- 5Approve one paid render. Ask for a styled image in plain language, review the cost card showing the tool, model, estimate and files to upload, then approve or decline.
What the skill makes the agent do
The skill lets the agent produce Photo to Anime work on the Beatra service. The styles listed in the skill are anime, manga, comic, cartoon, watercolor, clay and 3D character art. The agent is also instructed to review likeness and style consistency across a set of images instead of assuming they were preserved.
The catalog entry for this skill is a pointer to the installable package. It contains no client code of its own. The package it points to includes three Python scripts that make network calls and store a credential, so it is not a purely local tool.
- Turns a photo into a stylized character image using the Beatra service
- Reviews likeness and style consistency across a set of outputs
- Shows a cost estimate and waits for your explicit approval before any paid call
- Reports the delivered result and the credits charged
What you need before you install
You need four things: Git, to fetch the pinned package; Python 3.10 or newer, which the scripts use with only the standard library; a Beatra account with credits; and a browser, because signing in to Beatra opens one.
The skill is a poor fit if you need local-only or private image editing. Your uploads go to a third-party paid server, mcp.beatra.ai. Only continue if you are comfortable with that and can review the scripts.
- Git
- Python 3.10 or newer (standard library only, no dependency installation)
- A Beatra account with credits
- A browser for sign-in
How to install the skill
Download the ZIP from the directory page and unzip it into ~/.claude/skills/ for all your Claude Code projects, or into a project's .claude/skills/ folder for one project only. In the Claude apps, you can instead upload the skill under Settings, then Capabilities, then Skills.
The skill's own instructions describe a stricter route, and you should follow it if you can. Fetch the pinned package with Git, confirm the expected tree ID prints TREE-OK, inspect the files, and only then copy the reviewed folder to ~/.claude/skills/ai-photo-restyler. The instructions ask for a separate approval before the copy and another before authorizing, and they say to start a new agent session if your host finds skills only at startup.
- Read every file in the package before activation, including each script, reference and hook
- Do not treat a checksum or digest as proof of safety
- If you cannot complete the review, stop and ask for a separate review
dest="$HOME/.claude/skills/ai-photo-restyler"
test ! -e "$dest" && test ! -L "$dest" || { echo "destination exists; stop and ask the user"; exit 1; }
cp -R "src/skills/ai-photo-restyler" "$dest"How to trigger the skill with natural requests
You do not need special syntax. Ask in plain language and say clearly that you want the work done on Beatra and that you accept it is paid. The skill is meant for explicit requests, so vague wording such as improve this picture may not be enough.
After you ask, the agent should first show a cost card for one paid call. It lists the tool, the model, the estimate and which files will be uploaded. Nothing is uploaded until you approve that exact card. Planning or saying just do it does not count as approval.
- Turn this photo of me into anime-style character art on Beatra. It's paid, so show me the cost first.
- Make a watercolor version of this portrait using Beatra. Only upload the one file I name.
- Check that the Beatra connection works before we spend any credits.
Safety notes: scripts and network
The package runs three Python scripts: authorize.py, mcp_client.py and uninstall.py. Authorization opens a browser sign-in and stores a bearer token. The package has no self-updater, so it does not download code or change its own files. An update check only reports what the store or tool that installed the copy offers.
Some behavior is worth knowing before you approve. The free connection check is the safest first command. The uninstall script, shown below, removes the Beatra token files under ~/.beatra when it is the last Beatra skill on the device, but it never deletes the package directory itself.
- Network: calls go to mcp.beatra.ai
- Credential: a bearer token is stored after sign-in
- Local state: files are kept under ~/.beatra
- Paid calls only after explicit approval of a cost card
- Pinned Git tree: the steps check the expected tree ID before copying
python3 "$HOME/.claude/skills/ai-photo-restyler/scripts/mcp_client.py" verifyTips for using it well
Start with the free connection check before any render. Keep uploads to the files you actually want processed, because the skill is instructed to upload only images you supply and approve for the task. If a response is lost, the instructions say to retry only with the same request id and identical arguments, and never to submit a replacement while a task is still running.
To uninstall, run the dry run first. The output is a JSON decision: disconnected means this was the last Beatra skill on the device, keep_connection means another Beatra skill still uses the shared token, and revoke_retry means the server refused revocation and nothing changed, so run it again shortly.
- Run the free verify check first
- Upload only the images you choose
- Review each output with the agent before asking for another paid stage
- Use the --dry-run uninstall option before a real uninstall
dest="$HOME/.claude/skills/ai-photo-restyler"
python3 "$dest/scripts/uninstall.py" --dry-run
python3 "$dest/scripts/uninstall.py"What people use it for
- Turning a personal portrait into anime or manga style character art for a creative project
- Making a watercolor or comic-style version of a photo to compare looks before choosing one
- Producing a small set of stylized character images and checking likeness consistency across them
- Trying clay or 3D character styles for a character design sketchbook
Frequently asked questions
Is the Photo to Anime Restyler skill free?
The skill's free part is the connection check and the discovery calls, which the instructions describe as non-billable. Rendering is paid: each render is charged in Beatra credits, and the agent reports the net credits charged after each call. Check your Beatra account for the current prices, since this guide does not list them.
What do I need to run the Photo to Anime skill?
You need Git to fetch the pinned package, Python 3.10 or newer using only the standard library, a Beatra account with credits, and a browser for sign-in. The skill runs its scripts on your computer but does the image work on the Beatra cloud service.
Where do I install an agent skill like this?
For Claude Code, put the skill folder in ~/.claude/skills/ for all projects or in a project's .claude/skills/ folder for one project. In the Claude apps, you can upload the skill under Settings, then Capabilities, then Skills.
Are my photos private?
No. The skill sends uploads and prompts to the third-party paid service at mcp.beatra.ai. Only upload images you have agreed to send, and do not use this skill for local-only or private editing.
How is this different from editing photos locally in an app?
A local photo editor processes images on your own machine. This skill sends your chosen images to a hosted paid service and reports credits charged for each render. It also needs an agent that can run scripts, plus a Beatra account.
How do I remove the skill?
Run the uninstall script with the dry-run option first, then run it without that option. It removes the Beatra token files under ~/.beatra when this is the last Beatra skill on the device, but it does not delete the package folder, which you remove yourself.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you want to pay per render to turn your own photos into stylized character art and you are comfortable sending those photos to a third-party hosted service. Pick something else if you need local-only or private image editing, or if you cannot review Python scripts and an install process, because this package makes network calls and stores a bearer token.
Download the skill