DeepSelect

deepseek-recipe is a collection of Rust libraries and Python bindings for converting API requests in Messages, Chat Completions, and Responses formats into a shared conversation representation, encoding conversations into prompts or token IDs for DeepSeek V4 and V4.1 models, and converting model output back into protocol-specific responses. It supports streaming, text and image inputs, thinking and reasoning settings, function tools, tool calls, JSON object output, and stop sequences; its image component provides DeepSeek V4.1 preprocessing with OpenCV. The package parses model output and handles prompt encoding, while inference, tool execution, and HTTP transport are supplied externally. It includes separate components for core conversation types, prompt encoding, image processing, Python bindings, and example Axum and FastAPI servers with mock inference. Project code and public documentation are licensed under the MIT License.

DeepJIT is a lightweight, header-only C++20 library for just-in-time compilation, caching, loading, and launching of device kernels on NVIDIA CUDA GPUs and Huawei Ascend NPUs. It provides a shared runtime interface for both backends while keeping kernel source and compiler options backend-specific. The runtime hashes kernel source, tracked includes, compiler information, effective options, and application-provided dependency signatures to reuse compiled artifacts through in-memory and on-disk caches. Cache directories can be shared across users, processes, and nodes on local or distributed filesystems, with completed entries published through atomic directory renames. Lazy initialization defers device and compiler discovery until first use. DeepJIT integrates with PyTorch CUDA and torch_npu streams and can expose a runtime through pybind11. The CUDA backend uses NVCC and the CUDA Driver API to produce, load, and launch CUBIN files; the Ascend backend uses Bisheng, ld.lld, and ACL to compile, link, load, and launch kernels. It supports per-kernel compiler and launch options, compilation diagnostics, and optional PTX, SASS, or Ascend assembly dumps. The host environment requires Linux, a C++20 compiler with std::format support, Python, pybind11, and the dependencies for the selected backend; it is intended to be embedded in another C++ or Python extension as a header-only dependency.

DwarfStar is a self-contained, model-specific native inference engine for running DeepSeek V4 Flash, GLM 5.2, and, on high-memory systems, DeepSeek V4 PRO locally. It combines model loading, prompt rendering, tool calls, persistent KV state, an HTTP server, and a coding agent rather than serving as a general GGUF runner. The engine supports Metal on Macs, CUDA including multi-GPU systems, and ROCm on Strix Halo hardware; it also provides SSD streaming for machines with insufficient RAM, distributed tensor and pipeline parallelism, and server-side micro-batching for multi-user inference. The repository includes tooling and data for GGUF, imatrix, quality, and speed testing, and acknowledges code and design contributions from llama.cpp and GGML.

DeepSeek Harness (dsh) is an open-source agent harness developed by DeepSeek AI. It uses an architecture in which everything is a plugin and is powered by Cordis, whose design is described in “A Programming Paradigm for Spatiotemporal Composability.” The npm package can launch a local Web UI with `npx @deepseek-ai/dsh web`, serving at `http://127.0.0.1:3080` by default and opening the default browser; it can also be built and run from a repository checkout. The project is in developer preview and warns that compatibility-breaking changes may occur. It is licensed under the MIT License, with third-party dependencies and their licenses documented separately.

DeepSeek and Destroy (DSD) is a long-horizon coding-agent skill for executing large implementation plans through persistent worker loops. A premium parent agent handles architecture, decomposition, decisions, and phase gates, while fresh lower-cost specialist workers implement, review, repair, and recheck repository-scale changes. The default worker backend is external OpenCode using opencode-go/deepseek-v4-flash, and the skill is designed for Codex, Claude Code, OpenCode, Kilo Code, and comparable coding harnesses. Each task is rendered into an immutable JSON contract and managed through attempt directories containing prompts, reports, logs, state, and scope baselines. Its normal loop uses an Implementer, a fresh independent Reviewer, and—when needed—a fresh Fixer followed by another Reviewer. Python-based gates check objective integrity such as hashes, lifecycle, and exact source movement; an optional read-only Evidence Clerk interprets already-produced evidence for the parent but cannot invent proof, rerun verification, repair code, waive integrity failures, approve work, or recurse into another Clerk. Documentation is split between always-relevant parent rules and cold-loaded workspace, harness, transport, compaction, prompt, and specialist-role materials.