Claude Plays Tockers Trials — Development skill for Claude Code
A Claude Code skill that autonomously clears TFT's Tocker's Trials (PvE) using only screenshots and mouse/keyboard input, with rules distilled from its own mistakes.
How to install Claude Plays Tockers Trials
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open ShinkaUwU/claude-plays-tockers-trials and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Claude Plays Tockers Trials does
A Claude Code skill that autonomously clears TFT's Tocker's Trials (PvE) using only screenshots and mouse/keyboard input, with rules distilled from its own mistakes. Research demo only: automation violates Riot's ToS. Never use in PvP; PvE use may also get you banned. | Claude 自主通关云顶之弈「发条鸟的试炼」
Alternatives in Development
- Om Intake — Process all unread meeting notes in work/meetings/ — reads each file, classifies content, routes to the right 4.6k ★
- Native Feel Skill — An Agent Skill for designing cross-platform desktop apps that feel native — distilled from Raycast's 2.0 deep 1.9k ★
- Continuous Claude — Run Claude Code in a continuous loop, autonomously creating PRs, waiting for checks, and merging 1.3k ★
README
tft-play — a Claude Code skill that plays TFT Tocker's Trials
[中文说明见下方](#中文说明)
A [Claude Code](https://claude.com/claude-code) skill that lets Claude play the **Teamfight Tactics Set 18 "Tocker's Trials" PvE mode** on Windows: it reads the game from screenshots (computer use) and sends mouse/keyboard input through a small Python script.
⚠️ Read this first: account risk
**Riot Games' Terms of Service prohibit automation programs.** Section 7.1 of the [Riot Terms of Service](https://www.riotgames.com/en/terms-of-service) lists, among prohibited conduct, using "unauthorized third party programs, including mods, hacks, cheats, scripts, bots, trainers and automation programs that interact with the Riot Services in any way". Riot's [third-party applications policy](https://support.riotgames.com/en-us/riot/events/third-party-applications) applies to Teamfight Tactics, names "taking actions on your behalf (botting or scripting)" as not allowed, and says penalties range up to permanent suspension. Penalties under the Terms include account suspension or termination and hardware bans. The China server (operated by Tencent) also bans scripts that play the game for you.
This project is exactly such an automation program. Therefore:
- Never use it in matches against real players (Normal, Ranked, Hyper Roll, Double Up, or any other PvP queue). That would hurt other players and is the clearest kind of violation.
- Even in the PvE mode it was written for, using it is a risk. Neither Riot's Terms nor its third-party policy makes an exception for PvE, co-op or bot modes. Your account may be suspended or banned.
- If you run it, use an account you are prepared to lose, at your own risk. The author takes no responsibility for any penalty.
The project is shared for research and education on AI agents and computer use: how a model can plan, act and learn rules from its own mistakes in a real-time desktop app. It is not a tool for gaining an advantage
Related Skills
Omni Computer Use
Windows computer-use MCP server - 1:1 replica of Claude Desktop's tool surface, for the Claude Code CLI or any
Disk Cleanup
Diagnose disk usage and reclaim space autonomously — finds the biggest offenders by reclaim value and clears t
Hover Play
Hover play. A video card holds its poster until you point at it, then plays; leave and it pauses on the frame
Phone Remote
Turn your phone into a wireless keyboard + CLI launcher for a Mac. One tap opens Claude Code / Codex / any CLI
Lazy Bird
Universal development automation for ANY project. Claude Code implements features, runs tests, creates PRs aut
Aos
aos — one CLI over the machine that agents drive the same way you do. Install the agent skill, or serve it ove
Related Agents
Shux Tui QA
General hard-gate QA subagent for shux TUI, terminal UX, CLI, pane automation, plugin, attach, mouse/input, an
UX Architect.Agent
Use when designing or reviewing UX for the target UI repo: information architecture, navigation, user flows, i
Governance Learner
Feedback Distiller. Use this agent at sprint close to audit /memory/ logs, distill systemic patterns from bug-