Pithelo

Five agent repos gaining momentum. Every one measured.

Warming up: ranked by stars per day since creation.

Rank 1

lexmount/moli

Moli is a headless browser written in Rust for AI agents, with on-demand layout and rendering and a lightweight resource footprint. It fetches and extracts pages (Markdown, semantic tree, PNG, PDF), and exposes CDP, WebDriver Classic, and WebDriver BiDi through a CLI and an automation server that Playwright can connect to.

Developers building agents that fetch, search, or automate the web and want a browser runtime with model-friendly output formats and standard automation protocols should look at it.

  • ★ 19,984
  • Apache-2.0
  • README ≈ 6,707 tokens to load into context

AI-written summary

Score100.0
Rank 2

QingYunA/answer-me-with-html

An agent skill that lets the model write a short Markdown draft, which a bundled Node.js CLI turns into a single offline HTML page with generated diagrams and CSS. It also supports explainer videos and installs as a Claude Code plugin, a DeepSeek Harness plugin, or via `npx skills add` for other agents.

Worth a look for developers who want agents to answer complex questions with a readable page while cutting output tokens (the README reports about 612 versus 4,893 for hand-written HTML in its own benchmark, plus faster responses).

  • ★ 4,201
  • MIT
  • README ≈ 7,684 tokens to load into context

AI-written summary

Score100.0
Rank 3

kaankiziltug/logo-design-skill

A skill package for Claude, Gemini CLI, Codex and other Agent Skills-compatible agents that walks through logo design from brief to SVG delivery. It bundles design principles, dependency-free Python tools for SVG auditing, test sheets, mockups and favicon/app-icon export, and a searchable library of 1,400+ classified SVG logos for reference.

Developers building agents that need a structured, multi-step creative workflow with self-rendered image checks and a defined user checkpoint can use this as a worked example of the Agent Skills format.

  • ★ 4,044
  • MIT
  • README ≈ 13,054 tokens to load into context

AI-written summary

Score100.0
Rank 4

tigerless-labs/agent-memory

agent-memory is a long-term memory runtime for AI agents that stores memories as plain Markdown files, with a rebuildable local SQLite index (BM25 ranking, optional vector plugin) and a separate sleep-time Manage layer that consolidates and proposes deletions. Claude Code, Codex CLI, Muse Code, and other shell-capable agents share one store through a `mem` CLI, and it needs no API key.

Developers building agents that must remember across sessions, and who want inspectable, git-able memory files instead of an opaque vector or graph database, should look at it.

  • ★ 4,365
  • MIT
  • README ≈ 11,713 tokens to load into context

AI-written summary

Score85.6
Rank 5

Leonxlnx/unlazy

Anti-laziness skill for AI agents. Core: the Depth Tree method, which splits a task N layers deep and gives every leaf the full time budget of the whole task, so effort multiplies with depth. Grounded in 2025-2026 research on model laziness, underthinking and premature completion.

  • ★ 4,048
  • MIT
  • README ≈ 6,427 tokens to load into context

AI-written summary

Score54.7