fynnfluegge/agtx

DirSkills catalogs 7 skills from this repository, across 3 categories: AI Engineering, Automation, Quality.

1.2K stars111 forksView on GitHub
✅
2026/08/20

AGTX Execute

AGTX Execute implements an approved implementation plan in the current working directory, runs tests to verify changes, and writes a summary to .agtx/execute.md before stopping for further instructions.
Automation
1.2K111
🔍
2026/08/20

AGTX Research

AGTX Research explores the codebase to understand a task before planning, and writes findings to .agtx/research.md without modifying files. Use it during the research phase of an agtx-managed task to identify relevant files, architecture, complexity, and open questions.
AI Engineering
1.2K111
🔍
2026/08/20

AGTX Review

AGTX Review examines completed work for correctness, edge cases, and code quality, then writes findings to .agtx/review.md and stops.
Quality
1.2K111
📋
2026/08/20

Agtx Plan

Agtx Plan analyzes a codebase and writes a detailed implementation plan to .agtx/plan.md, then stops for user approval before any changes are made.
AI Engineering
1.2K111
🧹
2026/08/20

Agtx Sweep

Agtx Sweep extracts actionable work items from a coding conversation and creates them as tasks on an agtx kanban board. Use it when you want to capture, decompose, or hand off conversation results to agtx for execution.
Automation
1.2K111
📊
2026/08/20

Benchmark

Benchmark runs SWE-bench Lite evaluations against agtx coding agent workflows. Use it to configure, execute, and report on benchmark runs for Claude Code, Gemini CLI, or Codex CLI.
AI Engineering
1.2K111
💡
2026/08/20

Brainstorm Mode

Brainstorm Mode helps unpack a feature or enhancement idea through discussion only, without planning or implementation. Use it when you want to explore problems, trade-offs, and unknowns before handing off to the agtx board.
AI Engineering
1.2K111