Tag
Explore by tags
Compresses any context sent to LLMs (tool outputs, DB reads, RAG results, files, logs) to cut tokens by ~70–95% while preserving reversible originals; runs as a proxy or Python/TypeScript SDK with integrations for common agent frameworks.
Generates low-latency, streaming text-to-speech entirely on CPUs (no GPU or cloud API required), using an ~100M-parameter model with voice cloning and multilingual support. Optimized for low resource use (2 CPU cores, ~200ms to first audio chunk) — suited for local, privacy-sensitive, or embedded TTS.
Provides a framework to build, evaluate, and run AI SRE agents that investigate and remediate production incidents on your infrastructure. Includes a CLI, synthetic + end-to-end benchmark suites, and 40+ connectors for observability, infra, and LLM providers so teams can train agents and run investigations locally or in cloud.
Embeds Copilot's agentic workflows into applications via multi-language SDKs (TypeScript, Python, Go, .NET, Java, Rust). Uses the Copilot CLI runtime to handle planning, tool invocation, and file edits; supports BYOK and multiple auth methods.
Grows a personal skill tree by crystallizing each solved task into reusable skills; a ~3K-line autonomous agent framework that gives an LLM system-level control of browser, terminal, filesystem, input/vision and mobile (ADB) via nine atomic tools, optimized for low token cost.
Provides an operator-grade system and reusable toolkit for building, running, and securing agentic workflows across multiple LLM harnesses — skills, hooks, memory, MCP integrations, and security scanning for production agent deployments.
CLI for creating, running, and managing coding agents across local hosts, containers, and cloud sandboxes. Uses SSH, git, and tmux; supports snapshots, push/pull, auto-shutdown for cost control, and provider-agnostic workflows for developer-centric agent orchestration.
Large-scale mathematical reasoning dataset of model-generated solution trajectories produced with and without Python Tool-Integrated Reasoning (TIR), with final answers verified against reference solutions. Contains ~3.64M JSONL training samples (~144 GB) and per-source CC-BY / CC-BY-SA licensing; intended for training and evaluating tool-augmented mathematical reasoning in LLMs.
Management system for Xianyu sellers that handles multi-account operations, context-aware AI auto-replies, automated shipping/confirmation, and a web admin UI. Built with FastAPI, Playwright and Docker; intended for learning/research only, not for commercial use.
Runs background coding agents in isolated sandboxes to autonomously handle development tasks, create pull requests, and integrate with Slack, GitHub, Linear and webhooks. Supports multiplayer sessions, multiple LLM providers, fast startup via snapshots and prebuilt images; designed for single-tenant deployments.