Add PQS (Prompt Quality Score) to LLMOps

[OnChainAIIntel]

What this adds

PQS (Prompt Quality Score) — a pre-inference prompt scoring system, added to the LLMOps section.

Why it fits

PQS is a pre-inference LLMOps check. It sits in the production pipeline to gate prompt quality before any model call is made, saving LLM inference costs on low-quality prompts and preventing quality regressions from reaching production.

Closest neighbor in this list: Promptfoo (tests prompt outputs against assertions). PQS is complementary — it scores prompt inputs on 8 structured dimensions before inference runs.

What PQS does

• Scores any prompt on 8 dimensions: clarity, specificity, context, constraints, output format, role definition, examples, CoT structure
• Built on PEEM, RAGAS, MT-Bench, G-Eval, and ROUGE frameworks
• Published on GitHub Marketplace as "PQS Check"
• Alphabetical placement in table (P—between Portkey and PromptDX area)

Ecosystem

• GitHub Action: OnChainAIIntel/pqs-action
• Python SDK: prompt-quality-score (CrewAI, LangChain, AutoGen integrations)
• MCP server: pqs-mcp-server (Smithery 88/100, Glama Official)
• REST API: pqs.onchainintel.net (free tier, no signup)

MIT licensed. Happy to rework description or move to a different section if a better fit exists.