promptops

Prompt-as-Code for Enterprise AI. Standardize, audit, and deploy instructions across any AI coding assistant.

The best way to run CLI AI coding agents. Launch, orchestrate and monitor Claude Code, Codex, Gemini CLI and more in parallel — native desktop app.

Open-source backend foundation for multi-tenant AI SaaS: auth, RBAC, billing, AI gateway, PromptOps, RAG, API keys, webhooks, usage and audit. 开源多租户AI SaaS商业底座。

面向贷后催收的 LLM + RAG + 多智能体微服务系统（LangGraph / MCP / RocketMQ / FastAPI）。A production-grade LLM + RAG + multi-agent microservices reference for debt collection.

PromptOps framework treating prompts as engineering assets — lifecycle, versioning, context management, guardrails.

BFSI & AI LITERACY HINGLISH KNOWLEDGE GRAPH™ HINGLISH COGNITIVE ANCHORING MODEL™

Archived source - canonical development moved to PromptOps under packages/consensus-engine.

Production-grade LLM pipeline for viral content generation featuring PromptOps, Automated Evaluation (DeepEval), and Tiered Semantic Caching.

Archived source - canonical development moved to PromptOps under packages/eval-dataset-builder.

Trello for PromptOps: the cards of your boards on the Teams board, lists as columns. Runs in the plugin sandbox.

Archived source - canonical development moved to PromptOps under packages/prompt-package-manager.

Archived source - canonical development moved to PromptOps under packages/prompt-regression.

Archived source - canonical development moved to PromptOps under packages/model-scorecard.

Archived source - canonical development moved to PromptOps under packages/answer-diff.

One stop promptOps platform to track, optimize and store your agentic prompts

AI Agent Evals for PromptOps: how fast your AI models really answer, against their expected range and your own history. Runs in the plugin sandbox.

Reference architecture for prompt engineering systems as enterprise AI runtime infrastructure.

Fork for prompt lifecycle management, LLM observability, evals, prompt versions, datasets, and PromptOps patterns.

Archived source - canonical development moved to PromptOps under packages/benchmark-run-recorder.