Control the real Slay the Spire 2 game through commands connected to a local mod. Agents can read the current game state, play cards, choose routes, collect rewards, and manage shop and menu actions.
한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.
Command-line interface for Slay the Spire 2 - Control the real game through a local bridge mod HTTP API. Reads normalized game state and sends action commands for combat, navigation, rewards, and menu management.
설치하기 전에 에이전트에게 전달되는 지시 내용을 확인할 수 있습니다.
A stateful command-line interface for controlling the real Slay the Spire 2
game through the local STS2_Bridge mod. The CLI reads normalized game state
and sends action commands via a local HTTP API at localhost:15526.
This CLI requires a bridge mod that runs inside the game process. Both the CLI and the bridge are distributed from the same repository:
https://github.com/HKUDS/CLI-Anything
git clone https://github.com/HKUDS/CLI-Anything.git
cd CLI-Anything/slay_the_spire_ii/agent-harness
pip install -e .
The bridge mod is a .NET 9 plugin that must be compiled and installed into
the game directory. Full instructions are in the repository README, but the
short version is:
cd CLI-Anything/slay_the_spire_ii/agent-harness/bridge/plugin
./build.sh
cd ../install
./install_bridge.sh
If the build script cannot auto-detect the game data directory, set
STS2_GAME_DATA_DIR explicitly:
STS2_GAME_DATA_DIR="/path/to/data_sts2" ./build.sh
Launch Slay the Spire 2 via Steam, enable the STS2_Bridge mod in the mod
manager, then verify the connection:
cli-anything-sts2 state
If this returns JSON, the CLI and bridge are connected.
Prerequisites:
STS2_Bridge mod enabled.NET 9 SDK (only needed to build the bridge mod)# Read normalized game state (always start here)
cli-anything-sts2 state
# Start interactive REPL mode (default)
cli-anything-sts2
# Show all available commands
cli-anything-sts2 --help
| Command | Description |
|---|---|
state | Normalized state with decision field |
raw-state | Raw bridge JSON |
| Command | Description |
|---|---|
continue-game | Continue a saved run |
start-game --character IRONCLAD --ascension 0 | Start a new run |
abandon-game | Abandon the current save |
return-to-main-menu | Return to menu from any screen |
Characters: IRONCLAD, SILENT, DEFECT, NECROBINDER, REGENT
| Command | Description |
|---|---|
play-card <index> [--target <enemy_id>] | Play a card from hand |
use-potion <slot> [--target <enemy_id>] | Use a potion |
end-turn | End the current turn |
| Command | Description |
|---|---|
choose-map <index> | Select a map node |
proceed | Leave the current room |
| Command | Description |
|---|---|
claim-reward <index> | Claim a combat reward |
pick-card-reward <index> | Pick a card reward |
skip-card-reward | Skip the card reward |
claim-treasure-relic <index> | Claim a treasure relic |
select-relic <index> | Select a relic |
skip-relic-selection | Skip relic selection |
| Command | Description |
|---|---|
event <index> | Choose an event option |
advance-dialogue | Advance dialogue-only events |
rest <index> | Choose a campfire action |
| Command | Description |
|---|---|
shop-buy <index> | Buy from the shop |
| Command | Description |
|---|---|
select-card <index> | Select a card in overlay |
confirm-selection | Confirm the current selection |
cancel-selection | Cancel the current selection |
combat-select-card <index> | Select a card during combat overlay |
combat-confirm-selection | Confirm combat card selection |
| Command | Description |
|---|---|
action <name> --kv key=value | Send a raw bridge action |
The state command returns JSON with a decision field indicating the current
game screen. Route your next command based on this value:
| Decision | Meaning | Typical Next Commands |
|---|---|---|
menu | Main menu | continue-game, start-game |
combat_play | In combat, your turn | play-card, use-potion, end-turn |
hand_select | Card selection overlay | combat-select-card, combat-confirm-selection |
map_select | Map node selection | choose-map |
game_over | Run ended | return-to-main-menu |
combat_rewards | Post-combat rewards | claim-reward, proceed |
card_reward | Card reward pick | pick-card-reward, skip-card-reward |
event_choice | Event screen | event, advance-dialogue |
rest_site | Campfire | rest |
shop | Shop screen | shop-buy, proceed |
card_select | Card selection screen | select-card, confirm-selection |
relic_select | Relic selection | select-relic, skip-relic-selection |
treasure | Treasure room | claim-treasure-relic, proceed |
| Option | Default | Description |
|---|---|---|
--base-url | http://localhost:15526 | Bridge API URL |
--timeout | 10.0 | HTTP timeout in seconds |
state first to get the decision field아직 리뷰가 없습니다. 사용해 본 소감을 남겨 주세요.
로그인 없이 쓸 수 있습니다. 사용 후기, 질문, 작동하지 않은 환경 등을 편하게 남겨 주세요.
아직 댓글이 없습니다. 첫 댓글을 남겨 보세요.
Use Reasonix to build or improve a Python command-line interface for existing software. The workflow covers source analysis, interactive commands, JSON output, tests, documentation, and installation checks using the software’s own backend where possible.
한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.
Build and improve Python command interfaces for GUI applications with OpenClaw. The skill guides source analysis, command design, testing, and packaging, including interactive sessions and structured JSON output for agents.
한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.
Use Codex to build Python command-line tools for applications with graphical interfaces. Analyze source code, add commands, test against the actual software, validate packaging and documentation, and discover tools already generated or installed.
한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.
Inspect and edit 3MF files for 3D printing. Detect cylindrical holes, change their diameters, repair mesh issues, and compare files while preserving BambuStudio and PrusaSlicer metadata when repacking the edited file.
한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.
Manage AdGuard Home network filtering from the command line. Update blocklists, register devices, configure DNS rules that map names to addresses, and inspect query logs and statistics through the AdGuard Home REST API.
한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.
Generate slides, documents, websites, and diagrams from natural language prompts through AnyGen OpenAPI. Manage reference uploads, check generation progress, and download finished files using commands with structured output for AI agents.
한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.