balatro-mod-dev

Increment version and update changelogs

Health audit — check if project config, scripts, hooks, rules, file placement, and gitignore are up-to-date

Evaluate whether to compact context based on current workflow phase

Verify fix by extracting and analyzing Lovely logs after code changes

Draft a PR message for fork contributions

Get familiar with the current mod — what it does, how it works, key concepts

Fix transparent pixel colors in sprite images to prevent grey border artifacts during bilinear filtering.

List all available balatro-mod-dev commands

Initialize Balatro mod development for current repo

Review session work, surface new discoveries, and capture knowledge in the right place

Review codebase for refactoring opportunities

Publish a GitHub release with changelog and zip asset

Start sync with watch mode (run once at session start)

Run test scenarios and verify mod behavior via Lovely logs

Review and update all documentation — user docs, AGENTS.md — for accuracy, clarity, and freshness

Update plugin content based on new knowledge or instructions

Reviews code changes for correctness, edge cases, and quality. Use after writing code or before merging — catches bugs, logic errors, and mod-specific pitfalls. Only reports issues with confidence ≥ 80. <example> Context: Code was just written or modified user: 'Review the changes I just made' assistant: 'I will invoke the code-reviewer agent to analyze your changes.' <commentary>Triggers on review requests after code changes.</commentary> </example>

Executes implementation plans from the main agent. Use when the main agent has a concrete plan and needs Sonnet to write the code — multi-file edits, new modules, refactoring. <example> Context: Main agent has approved implementation plan user: 'Implement the approved plan' assistant: 'I will invoke the code-writer agent to execute the plan.' <commentary>Triggers after plan approval when code needs to be written.</commentary> </example>

Inspects Lovely runtime state — logs, dump metadata, and mod compatibility. Use when the user wants to check the latest log, analyze patch conflicts, or understand what mods changed in the game. Produces a structured diagnostic report for the main agent. <example> Context: User wants runtime diagnostics user: 'Check the latest log and game dump' assistant: 'I will invoke the debug-inspector agent to inspect runtime state.' <commentary>Triggers on log inspection, dump analysis, or mod compatibility checks.</commentary> </example>

Researches Balatro game source code to find function implementations, data structures, and injection points. Use when needing to understand how something works in the game, find where to inject code, or locate specific game mechanics. <example> Context: Main agent needs to understand game mechanics user: 'Find how G.GAME.blind works in the game source' assistant: 'I will invoke the game-source-researcher agent to find this in the game source.' <commentary>Triggers when game source research is needed.</commentary> </example>

Researches Lovely patch patterns and injection techniques. Use when needing to find how to patch game functions, understand Lovely syntax, or find examples of specific patch types. <example> Context: Main agent needs lovely patch syntax user: 'Find lovely patch examples for card evaluation' assistant: 'I will invoke the lovely-patch-researcher agent to find relevant patch patterns.' <commentary>Triggers when Lovely patch research is needed.</commentary> </example>

Researches how other Balatro mods implement specific features. Use when needing to see how existing mods solve a problem, find patterns for common functionality, or understand mod architecture approaches. <example> Context: Main agent needs to find implementation patterns user: 'How do other mods implement custom blinds?' assistant: 'I will invoke the mod-pattern-researcher agent to find patterns in installed mods.' <commentary>Triggers when mod pattern research is needed.</commentary> </example>

Extensive codebase exploration, understanding architecture, browsing files, token-intensive analysis. Use when the main agent needs a full map of the mod's structure. <example> Context: Main agent needs to understand project structure user: 'Map out this mod's architecture' assistant: 'I will invoke the project-explorer agent to analyze the codebase.' <commentary>Triggers when project architecture exploration is needed.</commentary> </example>

Synthesizes findings from multiple research agents into actionable conclusions. Use after gathering research from different sources — resolves contradictions, identifies the feasible approach, and produces a clear recommendation. <example> Context: Multiple research results need synthesis user: 'Analyze the research findings and recommend an approach' assistant: 'I will invoke the research-analyst agent to synthesize the findings.' <commentary>Triggers when multiple research results need cross-referencing.</commentary> </example>

Runs temporary scripts (Python, bash, Lua) and returns results. Use when main agent needs data from a one-off script execution - image processing, data extraction, format conversion - NOT when the script itself is the solution. <example> Context: Main agent needs data from a script user: 'Run this data extraction script' assistant: 'I will invoke the script-runner agent to execute and return results.' <commentary>Triggers when a temporary script needs to be executed for data.</commentary> </example>

Researches SMODS API patterns, hooks, and implementations. Use when needing to understand how to use SMODS features, find examples of SMODS usage in existing mods, or understand SMODS internals. <example> Context: Main agent needs to understand SMODS API user: 'How does SMODS.Joker registration work?' assistant: 'I will invoke the smods-api-researcher agent to find the SMODS API patterns.' <commentary>Triggers when SMODS API research is needed.</commentary> </example>

Plans implementation strategy for new features, refactoring, or structural changes. Use before implementing anything complex — produces a detailed plan with file changes, migration steps, risks, and alternatives. <example> Context: Complex feature or refactoring needed user: 'Plan how to implement this new feature' assistant: 'I will invoke the strategic-planner agent to design the implementation.' <commentary>Triggers on complex implementation requests.</commentary> </example>

Develop Balatro mods with Steamodded, Lovely, and SMODS. Includes game source navigation, mobile compat, and debugging.

9 hooks across 6 events