[NeurIPS 2026] Safety failures in coding agents and frontier LLMs. Reproducible workspaces for studying when AI agents generate unsafe data during task completion.
Repository Index
55,640 visible repositories in the index
121–144 of 203 · Most stars
🎬 Evidence-led Seedance 2.5 / 2.0 prompt library: 840+ cases traced to original posts, 490+ cross-model retest runs, 27 copy-ready templates, 79 installable AI-video Skills. Synced daily from goodcase.ai. EN / 中文 / 日本語.
- ai-video
- awesome
- bytedance
- prompt-engineering
Karate Club: An API Oriented Open-source Python Framework for Unsupervised Learning on Graphs (CIKM 2020)
- community-detection
- graph-clustering
- deepwalk
- networkx
Open-source, end-to-end platform for evaluating, observing, and improving LLM and AI agent applications. Tracing · Evals · Simulations · Datasets · Gateway · Guardrails. Self-hostable. Apache 2.0.
- ai-gateway
- observability
- simulation
- ai-agents
TranslatedLocal AI routing gateway with Agent client takeover — reuse existing client logins, call 50+ AI clients through one base URL, with unified OpenAI / Anthropic / Images APIs, multi-account routing, check-in, balance and request reports
- autoclaw
- catpaw
- local-proxy
- openai-compatible
A pre-execution guard for AI coding agents. It blocks destructive Git and file system commands, plus common attempts to access sensitive files, before a tool call runs. Supports Amp Code, Antigravity CLI, Claude Code, Codex, Cursor, DeepSeek Harness, Devin CLI, GitHub Copilot CLI, Grok Build, Hermes Agent, Kimi Code, OpenClaw, OpenCode, and Pi.
Translatedvm2api: converts subscriptions into an API using a fully realistic virtual machine emulation approach
A robust Python tool for text-based AI training and generation using GPT-2.
Open decision models from 0.8B to 27B: a calibrated probability for every option in one pass, and a short thought only when unsure. Engine for vLLM, llama.cpp and MLX.
- agents
- calibration
- classification
- decision-making
Local CLI Copilot, powered by Ollama. 💻🦙
A Go-based OpenCode Zen / Zen Go API proxy with OpenAI & Anthropic API compatibility, key/proxy pooling, automatic routing, and WebUI.
Aion 2 AI Prompt Engineering Toolkit - Standalone Desktop Trainer Setup
- aion-2
- aion-2-2026
- aion-2-game
- aion-2-optimizer
Simple shell script to use OpenAI's ChatGPT and DALL-E from the terminal. No Python or JS required.
- chatbot
- chatgpt
- chatgpt-api
- chatgpt-api-wrapper
GVS5H: Five Qwen3.8-27B Models Match Claude Fable 5 on LiveCodeBench Hard | Fable 5 Level Coding for a Fifth the Price - or on a Single GPU
TranslatedA Rust-based, low-risk-control, high-performance, self-hosted multi-account transparent proxy gateway
TranslatedOne-click replacement of the built-in prompts of four AI clients: DSH / WorkBuddy / ZCode / Codex. Double-click to use, self-verifiable, backs up before changes, one-click restore. A single-file exe that needs no Python installation, open source under the MIT license; activation codes are used for server operation and maintenance, and are issued through self-service on the official website.
- deepseek-harness
- persona
- python
- windows
TranslatedA multi-account gateway that turns Tencent WorkBuddy accounts into an OpenAI-compatible API, while automatically completing all tasks in the task center. Includes a web management panel (account pool visualization / points tasks / hot configuration updates). An enhanced fork based on Sliverkiss/workbuddy2api
This Word Does Not Exist
Run large AI models on your own PC: Qwen3.8-Flash-Next on Strata's engine, and GLM 5.3 Flash, DeepSeek V4 Flash, LFM2.5, Qwen3.6, Ornith 1.5, Gemma 4, Tiel Coder and MiniCPM-V on a second engine written for PolyStrata. One-click install for Windows and Linux, OpenAI/Anthropic API on localhost.
- anthropic-api
- cuda
- gemma
- gguf
🐋 Run 284B DeepSeek-V4-Flash (<59GiB RAM, 44.4 tok/s pp, 9.1-14.7 tok/s) & 30B MoEs (25-111 tok/s) on a pure 64GB CPU at 256K ctx (0 GPUs). SwiGLU-H2, Stockfish-PVS L3 & 13ms Claude Code restore.
Fast vectorization, topic modeling, distances and GloVe word embeddings in R.
- word2vec
- text-mining
- natural-language-processing
- glove
37 tok/s from Qwen3-Next-80B (48 GB) on a 32 GB laptop, 24 without drafts vs 2.8 for llama.cpp; reading 62 vs 16 tok/s. Faster than llama.cpp on all 13 models compared in Q4_K_M. LLMs bigger than your RAM, pure C, CPU or GPU.
- avx2
- c
- cpu-inference
- gguf
A simple deep learning library for estimating a set of tags and extracting semantic feature vectors from given illustrations.
OpenAmika: MekaHime's AI companion Amika, open source. Live2D avatar, real-time voice, long-term memory, OpenRouter or Ollama brain.
- ai-companion
- live2d
- long-term-memory
- mekahime