try-works

try-works / recursive-mode

Public

Recursive workflow for agentic engineering. Like Factory Missions but properly recursive, free and open source.

recursive-mode.devintroduction
17
2
100% credibility
Found Apr 13, 2026 at 17 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

recursive-mode provides a file-based workflow for AI agents to manage software development tasks through structured phases like requirements, planning, implementation, testing, review, and memory updates.

How It Works

1
🔍 Discover recursive-mode

You hear about a helpful tool that organizes AI assistants for building software step by step, keeping everything in neat notes instead of chat history.

2
📦 Add the skill

You easily add this organizing skill to your AI helper's toolkit so it can use structured steps for your projects.

3
🗂️ Open your project

You go to your project's main folder where your code lives.

4
🚀 Start a task

You simply tell your AI 'Start working on this task' and it creates a dedicated space with checklists for what needs to be done.

5
📝 Follow the guided steps

Your AI walks you through planning what to build, making changes safely, testing everything, checking the work, and wrapping up with lessons learned.

6
🔒 Review and finish

You check the clear records of each step, approve the results, and your project updates with proof everything was done right.

✅ Task complete

Your project is improved, all work is documented for easy review later, and your AI remembers tips for even better future tasks.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 17 to 17 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is recursive-mode?

recursive-mode is a Python package that structures AI agent workflows into file-backed runs for requirements, planning, implementation, testing, reviews, and closeout—ditching flaky chat histories for auditable repo artifacts. It bootstraps a .recursive/ directory with memory for lessons, state tracking, and decisions, plus subskills for isolated worktrees, TDD, debugging, review bundles, and subagents. Install via npx skills add try-works/recursive-mode, then kick off runs with commands like "Implement the run."

Why is it gaining traction?

Unlike loose recursive LLM prompts or n8n-style workflows, it enforces draft-audit-repair-lock cycles with git-verified diffs, subagent checks against real files, and durable memory that feeds future runs—cutting vague "done" claims. Developers hook on the recursion: phases revisit outputs, promote lessons, and verify delegations, making agentic engineering reliable without custom scripting. Short commands resume runs, blending github actions recursive workflow vibes with open-source freedom.

Who should use this?

AI agents building production code who need audit trails, teams delegating to subagents but verifying outputs, or solo devs running recursive LLM tasks like TDD cycles or review bundles in git repos. Ideal for Python projects where chat drift kills momentum, but skip if you prefer un opinionated tools.

Verdict

Early alpha with 17 stars and 1.0% credibility—docs shine but expect rough edges like manual bootstraps. Worth a spin for agent workflow tinkerers; fork and mature it if structured recursion clicks.

(187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.