GitHub 2026-06-11 Weekly Trending Projects

Last30Days is a multi-source trend briefing tool that searches and scores signals from social platforms, news, and prediction markets over the last 30 days, then synthesizes cited narratives—designed for researchers, prompt engineers, and product teams tracking short-term trends.

Headroom delivers a local, reversible context-compression layer that uses content routing and multiple specialized compressors to substantially reduce tokens sent to LLMs; suited for cross-agent shared memory and token-cost-sensitive production or research, but license and maintenance activity should be verified before adoption.

Open Notebook is a privacy-first open-source research notebook that provides multi-model integrations, multimodal content ingestion and vector search; it targets teams and developers who require data sovereignty, self-hosted deployment and advanced content generation such as multi-speaker podcasts.

This repository assembles curated Codex plugin examples and cross‑platform reference implementations that use a unified manifest and companion surface structure; it serves as a practical reference for plugin development, integration, and teaching, but prospective adopters should verify license and maintenance details to mitigate adoption risk.

Agent Reach bundles common web, social and video CLI tools into a one‑command scaffold so AI agents gain internet‑reading capabilities quickly, keeping upstream tools replaceable and minimizing integration effort.

Taste Skill supplies composable frontend 'skills' and an image-to-code pipeline that use adjustable design dials and GSAP scaffolds to accelerate turning AI outputs into implementations; however, missing license and contribution metadata require compliance and maintenance validation before production adoption.

MarkItDown converts varied documents into structured Markdown for LLM pipelines, enabling automated text workflows while requiring care with optional dependencies and plugin compatibility.

PM Skills Marketplace turns established product-management frameworks into callable skills and chained workflows; using assistants like Claude and Codex it operationalizes product decisions, suited for teams needing structured, repeatable PM processes.

Aimed at users who want local voice-interactive experiences with Live2D avatars, offering modular LLM/ASR/TTS integrations and desktop/web clients — but evaluate license and maintenance activity risks carefully.

Hermes Agent is a server-deployable autonomous AI agent offering a terminal-first interface, multi-platform messaging gateway, persistent memory and skills hub, sandboxed terminals, and scheduling—suitable for long-running, auditable, and extensible automation in engineering and operations.

Trivy is a lightweight Go-based security scanner that inspects images, filesystems, Kubernetes, and Git targets to produce SBOMs, CVE findings and IaC issues—designed for CI/CD and platform integration with broad ecosystem support.

ECC is a production-focused AI agent operator system offering cross-harness skill libraries, memory optimization, continuous learning, and security tooling—designed for engineering and research teams deploying agent orchestration at scale.

A production-oriented, embeddable framework for in-app AI assistants that offers fast CLI onboarding, customizable UI components and agent workflows—ideal for teams needing deep application integration.

Harness is a team-architecture factory for Claude Code that converts domain descriptions into six types of agent teams and matching skills, enabling rapid construction, orchestration, and validation of multi-role automated workflows.

Impeccable supplies an LLM-driven frontend design skillset—commands and references to audit, polish, and extract UI patterns—enabling teams to apply consistent design best practices across AI tools.

A Claude Code-focused compound-engineering plugin and CLI that defines cyclic workflows, exposes a plugin marketplace, and offers experimental OpenCode/Codex conversions; suitable for engineering teams with self-managed tooling to trial in controlled environments.

Supervision is an engineering-oriented Python CV toolkit offering model-agnostic connectors, customizable annotators, and a complete dataset toolchain to accelerate inference, visualization, and dataset workflows.