claude-skill

GenericAgent

Self-evolving agent: grows skill tree from 3.3K-line seed, achieving full system control with 6x less token consumption
14,122 stars1,648 forksPythonUpdated 9/3/2026100% free · open source
What it does

A self‑evolving Python agent that expands its own skill tree from a 3.3 K‑line seed and runs full‑system tasks while using six times fewer tokens than comparable Claude‑based agents.

When to use it
  • You need an autonomous AI worker that can learn new capabilities on‑the‑fly without rewriting prompts.
  • Your product requires low‑cost token usage for heavy‑weight Claude calls (e.g., data pipelines, automation scripts).
  • You want a single Python‑based agent that can take over OS‑level actions (file I/O, subprocesses) without building separate tool wrappers.
Ready-to-paste prompt
python -m generic_agent.main --seed ./seed/seed.py --task "Scrape the latest pricing data from competitor sites, store it in a PostgreSQL table, and email a summary report to the product team."
Heads up: The agent requires an Anthropic Claude API key (set in CLAUDE_API_KEY) and runs only on Python 3.10+; missing the key or using an older interpreter will cause immediate startup failures.
Saves to your device
Use with Claude
New

Skip the builder — one click puts this in Claude, Cursor, Antigravity and more.

⚡ Runs out of the box

A prompt/skill package — nothing to install beyond adding it to your AI tool.

Try it instantly — no install
Claude Code
mkdir -p ~/.claude/skills/genericagent && curl -fsSL https://workflowstacks.com/api/skills/genericagent/claude-skill -o ~/.claude/skills/genericagent/SKILL.md
Open in another AI app

Opens the app with this repo with the prompt ready to go — no copy-paste needed.

Connect the whole catalog (MCP)
claude mcp add --transport http workflowstacks https://workflowstacks.com/api/mcp

Adds a WorkflowStacks connector to Claude Code: search and load any skill here by chatting.

How GenericAgent works
Codeflow
Free to inspect

GenericAgent is a large Python project (~46k lines across 104 code files, plus 15 test files). It is a full software project: use it through its install path rather than reading it end to end. Last commit this month, MIT license, has a test suite.

Size
Large codebase
~46k lines · 104 code files · ~6 h to skim
Setup
Developer setup
A real software project. Use it via its install path; don't expect to read it all.
Runs on
Python
No API keys detected
Python 75%JavaScript 11%Rust 4%CSS 3%HTML 3%
Where to start reading
  1. 1
    README.md
    Start here — what it does and how to install it
  2. 2
    ga_cli/cli.py
    Where the program starts running
  3. 3
    pyproject.toml
    Dependencies and the commands it exposes
What's in each folder
plugins/Sub-packages / plugins4 files
docs/Documentation5 files
frontends/Folder145 files
memory/Folder43 files
ga_cli/Folder5 files
reflect/Folder5 files
assets/Images & static assets37 files
.github/CI / automation (GitHub Actions)2 files
READMEHas testsDocumentedCI checksMIT licenseUpdated this month
Quick Actions
Details
Creator
lsdefine
Language
Python
Category
claude-skill
Published
1/16/2026

Are you the creator of this tool? Claim your listing → and earn 85% of every sale.