STS2MCP AGENTS.md

STS2MCP AGENTS.md is an instructions file for Codex, OpenCode from Gennadiyev/STS2MCP. It costs 820 tokens per session, scanned A, original, MIT.

Gameplay instructions for controlling Slay the Spire 2 through an MCP connection, a tool interface that lets an agent read game state and perform actions.

In plain words
What is it for?
Use them when polling combat state, playing cards safely, choosing event or rest options, claiming rewards, and using potions through the available game tools.
Why use it?
They address tool-specific issues such as delayed state updates, changing card positions, target selection, and event or reward sequences.

Instructions file for CodexOpenCode

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.

agentmods
npx agentmods add instructions/gennadiyev/sts2mcp/agents-md
Clone the repo
git clone --depth 1 https://github.com/Gennadiyev/STS2MCP

Made for: Codex, OpenCode.

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 STS2MCP AGENTS.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/gennadiyev/sts2mcp/agents-md.svg)](https://agentmods.dev/instructions/gennadiyev/sts2mcp/agents-md)
Your own site
<a href="https://agentmods.dev/instructions/gennadiyev/sts2mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/gennadiyev/sts2mcp/agents-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 820 This file is loaded in full into every session.
When invoked 820 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
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 $0.00820 $0.00820
Opus 5 $0.00410 $0.00410
Sonnet 5 $0.00164 $0.00164
Haiku 4.5 $0.00082 $0.00082

Measured 5d ago against content hash d7c5b46ee3d3, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

STS2MCP AGENTS.md 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.

AGENTS.md · 62 lines

How it starts

The opening of the file, as written. The whole thing — 62 lines — stays where its author put it; the contents beside it link to each section on GitHub.

STS2 MCP — AI Gameplay Guide

MCP Tool Calling Tips

State Polling

  • After combat_end_turn, the state may show is_play_phase: false or turn: enemy. Call get_game_state again to advance to the next player turn.
  • Sometimes you need to call get_game_state twice — once to see enemy turn results, once to see your new hand.
  • Use format: "json" during combat for structured data; format: "markdown" for map/event overview.

Card Index Shifting

  • CRITICAL: Playing a card removes it from hand and shifts all indices. Play cards from RIGHT to LEFT (highest index first) to keep lower indices stable, or re-check state between plays.
  • When targeting, always provide target for single-target cards. Entity IDs are UPPER_SNAKE_CASE with a _0 suffix (e.g. KIN_PRIEST_0).

Event & Reward Flow

  • Events: event_choose_option. After choosing, there's often a "Proceed" option at index 0.
  • Rest sites: rest_choose_option, then proceed_to_map.
  • Rewards: claim from right-to-left (highest index first) to avoid index shifting. Card rewards open a sub-screen; use rewards_pick_card or rewards_skip_card.

Potions

  • use_potion(slot=N) — slot is the potion slot index, not a card index.
  • discard_potion(slot=N) — discard a potion to free up the slot when full.
  • Potions don't cost energy or count as card plays. Use buff potions BEFORE playing cards.

General Strategy

Core Principles

  1. HP is a resource, not a score. Take calculated damage to deal more. Don't waste energy on block when enemies aren't attacking.
  2. Deck quality > deck size. Skip card rewards if nothing synergizes. A lean deck draws key cards more often.
  3. Front-load damage. Killing enemies faster means less total damage taken.
  4. Read intents carefully. Sleep/Buff = go all-out offense. Attack = balance block and damage. Debuff = usually no damage, offense turn.

Combat Sequencing (General)

  1. Play 0-cost utility/setup cards first.
  2. Play skills before attacks when possible — many mechanics reward this order (e.g. Slow debuff on enemies stacks per card played).
  3. Play biggest attacks last to benefit from accumulated buffs/debuffs.
  4. Check enemy HP — if you can kill this turn, skip blocking entirely.

Read the full file on GitHub · 62 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. 5d ago First seen · 62 lines · 820 tokens per session scan A d7c5b46ee3d3

Subscribe to this mod's changes

STS2MCP AGENTS.md is an instructions file published in the GitHub repository Gennadiyev/STS2MCP (482 stars, last pushed 1mo ago), licensed MIT. It adds 820 tokens to every session, about $0.0041 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 2026-08-30.