Multi-agent collaboration for Claude Code via filesystem-mediated turn-taking. N agents on an arbitrary graph (Petersen, ring, star, pair) write turns to per-edge CONVO.md files with atomic .turn sentinels. Lossless DAG-based archive layer keeps long conversations cheap to re-read. Round 15d ships ephemeral-only…
★not rated 2 4mo agoA
tokens not measured
originalMIT
Opinionated skills for building and debugging LangGraph agents and for prompt engineering: pick a stack, write a spec, debug a broken graph, and write better prompts.
★not rated 2 2mo agoA
tokens not measured
originalMIT
First Principles thinking mode. Forces AI to break out of analogical reasoning and re-derive solutions from fundamental truths. Five-step framework: identify assumptions, strip them, return to ground truths, derive from scratch, compare.
★not rated 2 2mo agoA
tokens not measured
originalApache-2.0
Full-lifecycle engineering skills for AI coding agents — orchestration chain from brainstorm to ship, with production-grade engineering standards at every step.
A collection of command-line development skills for structured workflows, testing, code review, end-to-end checks, and shared team knowledge. TDD means writing tests before code; SDD refers to designing or specifying work before implementing it.
★not rated 3
changed todayA
tokens not measured
originalMIT
The autonomous engineering system for repository portfolios — engineer, read-only surveyor, and meta-engineer agents; portfolio, product, spend, and improvement workflows; cross-tool instruction architecture and skill discovery; configured by the consumer AGENTS.md.
★not rated 2
changed todayA
tokens not measured
originalApache-2.0
Source-first AI-agent toolkit with native Claude Code plugin metadata, English setup routing, Windows-safe hook repair, opt-in repo auto-update, and local bridge adapters.
slashwork offloader. Interception is on by default: /work init authenticates (a bare /work with no token does the same), and from then on a PreToolUse hook routes self-contained subagent spawns to the offload network and returns the artifact in place, falling back to the local spawn on any miss. /work off pauses…
★not rated 2 5d agoA
tokens not measured
originalMIT
Developer tools for building with the Gravity Rail SDK and CLI — guides, patterns, and recipes for workspace automation, AI assistants, and multi-channel communication.
Scheduler/launcher for Claude Code headless runs. Cron-scheduled jobs launch claude -p in a folder, with per-job model, reasoning effort, timeout, duplicate guard, missed-tick catch-up, log rotation, a process-level circuit breaker that treats a Claude usage-limit / transient API overload as a skip (not a failure) so.
Get a second opinion from Google's Gemini on code changes, plans, documentation, architecture, and other work products. Use when you need independent validation, cross-checking critical decisions, catching blind spots, or getting multiple AI perspectives on important work.
★not rated 2 8mo agoA
tokens not measured
originalMIT
Almost-autonomous agent harness that takes a GitHub issue or Linear story through planning, review, implementation, hardcore code review, and draft-PR creation. Self-tunes its subagents via a meta-reflection loop within hard safety boundaries.
★not rated 2 2mo agoA
tokens not measured
originalMIT
LmHeads A2A network plugin. Discover agents, invoke their skills, receive and respond to tasks. Skills CRUD + opaque delegation + end-to-end-encrypted vault for credential sharing. Run /lmheads:configure once to save your agent API key; the plugin reads it at boot.
Production-grade context engineering tool for expert LLM engineers. Analyzes and improves plugin architectures using SOTA orchestration patterns - Four Laws of Context Management, tiered fidelity, severity-batched grounding, and firewall architectures.
★not rated 2 4mo agoA
tokens not measured
originalBSD-3-Clause
Multi-phase agent orchestration for Claude Code. Inspired by Bun's migration workflow pattern. Bootstraps .claude/workflows/ in any project with TypeScript runtime, JSON Schema validation, and parallel agent execution.
Build MCP servers in Delphi with MCPConnect: create a new server (stdio, WebBroker or Indy), add tools, resources, templates, App UIs and prompts, and set up sessions, notifications and OAuth security.
★not rated 2 29d agoA
tokens not measured
originalMIT
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: