中/EN

Home

Open Source

Curated open-source AI projects with verified GitHub data and hands-on breakdowns.

Frontline HotspotHardcore ReviewsOpen SourceField SOP

Hand Your Phone to an Agent: phone-harness Hits 2,000 Stars in Two Weeks, and Your iPhone Can Be Fully AI-Driven Without a Jailbreak

ShawnPana/phone-harness (1,977 stars / 183 forks, MIT, Python; GitHub API snapshot 2026-08-22, created Aug 7 - two weeks old): let your agent control your phone. No iPhone jailbreak - macOS Sequoia+'s iPhone Mirroring window is the whole transport: screencapture plus Vision OCR for "the poor man's DOM," CGEvents for HID-level taps and keystrokes. Android goes over adb (USB/Wi-Fi) with screencap, the exact uiautomator accessibility tree, and input commands. Installation is a single prompt pasted into Claude Code or Codex, with a --doctor self-check. Philosophy: no prebuilt tools - whatever's missing, the model writes mid-task in ordinary Python (agent_helpers.py). Includes the learned-the-hard-way pitfall list (AppleScript clicks silently fail, unicode input impossible, slow drags don't move iOS lists) and three safety gates (spare phone / payment approval / API first).

Give an LLM an Entire Mac: browser-use Ships a "Zero Rails" Harness, 480 Stars in Two Days

browser-use/macos-harness (480 stars / 33 forks, MIT, Python; GitHub API snapshot 2026-08-19, created Aug 17 - two days old): the browser-use org's thinnest possible harness that "gives an LLM complete freedom" over a Mac - no framework, no recipes, no rails. Six primitives (see / key / type / click / ax / script) cover the whole machine: CGWindow screenshots of background windows without raising them, CGEvent keystrokes straight to a target PID, raw AX and Apple Events as fallback, plus CDP driving a real logged-in Chrome - with browser, Path, and subprocess ready in the same persistent Python process. The philosophy: no Spotify tools, no Slack tools - the model writes whatever logic is missing in ordinary Python, mid-task. Installation is a single prompt you paste into Codex or Claude Code (uv + Python 3.12 + skill registration + doctor permission self-check). doctor reports the exact macOS permissions needed; anonymous telemetry records command categories only and can be disabled with one command. Ecological niche versus browser-use (clean browser) and ego-lite (shared logged-in browser). Read against this week's OpenAI rogue-agent news: no-rails and guardrails are two sides of one coin. Experimental, macOS only.

AI Agents Want Group Chats Too: cumora by the avante Author Puts Claude Code and Codex on the Same Roster (1.4k Stars in One Day)

yetone/cumora (1,460 stars / 158 forks, MIT, TypeScript; created Aug 17, API same-day snapshot): the avante.nvim author's team chat for AI agents - agents as first-class teammates with personas, memory, atomic work claims, real email, and shared Kanban/calendar. Two brain paths: Cumora Cloud (per-agent K8s pods on the OpenAI Responses API) or BYOA (npx cumora agent computer - your local Claude Code/Codex on your own subscription; the server never sees keys). Anti-collision trio (seen-cursor freshness gate / atomic claims / small-brain triage), an llm_calls cost ledger, and a CI big-model guard. Five minutes local: Postgres + Redis + OPENAI_API_KEY. Five cautions include day-one maturity and cloud token burn.

The Watermark Army Goes Open Source: Inside Google's SynthID-Text Reference Implementation

google-deepmind/synthid-text (1,024 stars / 96 forks, Apache-2.0, API snapshot Aug 17): the official reference implementation of SynthID text watermarking, tied to a Nature paper. The watermark is woven into the sampling distribution (keys config + HF Transformers mix-in on Gemma/GPT-2), with dual detectors (training-free Weighted Mean / trainable Bayesian) and a runnable Colab (2B needs T4 / 7B needs A100). Officially research-only; the production version lives in Hugging Face Transformers; accumulate_hash offers no cryptographic guarantees. Three angles (adversarial/compliance/engineering) plus five cautions. Counterpart to watermarks-remover Layer B.

ai-manju: A Local Production Pipeline for AI Short Dramas - ComfyUI + MiniMax-H3 from Chinese Draft to Video with Audio

catiseyeqaq/ai-manju-shengcheng-xitong: a local production pipeline for realistic AI films/short dramas. ComfyUI 0.31.0 orchestration + MiniMax-H3 joint audio-video generation (BF16 weights ~129GB, 8x PPU-ZW810E) + Qwen3.6-35B-A3B Chinese prompt polishing. 33 template workflows, T2V/I2V/R2V, majicFlus + PuLID cross-shot face lock + H3 first/last-frame continuity + 48k audio alignment. Full-chain analysis, hardware bar, 5 FAQs. Early personal project (4 stars); verify via Releases.

GLM-5.3 Deep Dive: Zhipu's #1 Open-Source Coder That Also Finds Bugs - "Open Shield" Launched

Zhipu GLM-5.3 (2026-08-14): base unchanged, post-training scaling lifts the intelligence ceiling, making it the strongest open-source coding model (#1 open-source on Terminal Bench 3.0 etc.), with emergent cybersecurity capability (code review/vuln discovery) - 2436 vulnerabilities found with security teams, "Open Shield" program launched. Deep analysis + 5 pitfalls. Repo/stars per official; not a benchmark.

oh-my-openagent: The OpenCode Plugin That Got Anthropic to Block OpenCode -- Multi-Model Orchestration with a SUL License Catch

code-yeongyu/oh-my-openagent (67k stars, TypeScript, pushed today) is a batteries-included OpenCode plugin with multi-model orchestration, parallel background agents, and custom LSP/AST tools. Its README declares "Anthropic blocked OpenCode because of us -- Claude Code is a nice prison." Pitfall: licensed under SUL-1.0, not MIT, with usage restrictions.