A model-driven harness for AI coding agents: semantic modeling, app generation, progressive API Assist and evidence-based verification.
-
Updated
Sep 27, 2026 - Python
A model-driven harness for AI coding agents: semantic modeling, app generation, progressive API Assist and evidence-based verification.
AI-native Rust library and runtime for business apps: model-driven, typed SQL queries, safe expressions and audited graph mutations.
Rust + SQLite task board showcasing TeaQL: composable queries, business state transitions, SQL tracing and field-level audit trails.
AI-native Java library and runtime for business apps: model-driven, typed SQL queries, safe expressions and audited graph mutations.
TeaQL CLI for coding agents and developers: evaluate business models, generate apps and discover typed APIs with progressive Assist.
Maven plugin for TeaQL model-driven code generation: generate typed domain libraries and application workspaces from business models.
AI-native Go library and runtime for business apps: model-driven, typed SQL queries, safe expressions and audited graph mutations.
Store artifacts
AI-native Swift library and runtime for business apps: model-driven, typed SQL queries, safe expressions and audited graph mutations.
AI-native C#/.NET library and runtime for business apps: model-driven, typed SQL queries, safe expressions and audited graph mutations.
AI-native Python library and runtime for business apps: model-driven, typed SQL queries, safe expressions and audited graph mutations.
AI-native TypeScript library and runtime for business apps: model-driven, typed SQL queries, safe expressions and audited graph mutations.
Generated, typed, auditable business tools for Vercel AI SDK agents
TeaQL Coding Agent for NVIDIA DGX Spark — Ratatui TUI with 64K context budget
Study of complex business system benchmark in TeaQL Java and TeaQL Rust
Does reducing the solution space through semantic APIs, generated models, compile-time constraints, and runtime verification improve the reliability and efficiency of LLM-based software development?
Context lifecycle tools for coding agents: consume-once messages and explicitly scoped instruction blocks, with host-specific adapters.
To associate your repository with the teaql topic, visit your repo's landing page and select "manage topics."