An AST-First transpiler enabling AI agents to write structured JSON AST directly, compiling it into type-safe Go, Rust, TypeScript, Kotlin, Swift, and Python.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install xiaoqinli
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install xiaoqinli using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Xiaoqinli (XQL)?
Xiaoqinli (XQL) is an AST-First compiler engineered specifically for AI agents, bypassing the conventional source-parsing paradigm. Rather than forcing LLMs to output complex syntax-heavy source files that are prone to compilation and parsing errors, agents write structured .xql.json files (JSON Abstract Syntax Trees) directly. The compiler validates types, effects, and capabilities at compile time, eliminating syntactic mistakes and ensuring high structural integrity prior to generating clean, idiomatic source code in six target languages.
Using this capability within your toolkit of Openclaw Skills ensures that code generated by AI agents is structurally correct, type-safe, and secure. Developers can program AI agents to output robust systems in Go, Rust, TypeScript, Kotlin, Swift, or Python, with the compiler handling the translation and safety checks transparently.
Xiaoqinli (XQL) Use Cases
- Syntax-Free AI Generation: Enable LLM agents to write logic directly into JSON Abstract Syntax Trees to eliminate syntax mistakes.
- Multi-Language Software Synthesis: Transpile a single abstract declaration file into equivalent code for Go, Rust, TypeScript, Kotlin, Swift, or Python.
- Static Guarantee and Safety Enforcement: Verify capability grants and computational side effects of generated code before compiling to native binaries.
- Automated Code Pipeline Integration: Standardize AI agent compilation workflows using Openclaw Skills in continuous integration and deployment environments.
How Xiaoqinli (XQL) Works
- AST Construction: The AI agent produces program structures using a predefined JSON schema with defined nodes (Declarations, Statements, Expressions).
- Type Checking: The compilation system evaluates function signatures, return types, variable declarations, and mapping rules to assign compile-time type safety.
- Effect Inference: The compiler computes transitive side effects, ensuring functions declaring pure behaviors cannot execute impure operations.
- Capability Checking: The transpiler verifies caller permission scopes, ensuring they encapsulate callee requirements (e.g., IO, filesystem, network access).
- Target Code Emission: Validated trees are transpiled to clean, idiomatic syntax in Go, Rust, TypeScript, Kotlin, Swift, or Python.
Xiaoqinli (XQL) Setup
Prerequisites
- Go 1.23+ is required to build from source.
Installation
# Clone and build the transpiler binary
git clone https://github.com/Freecode100Year/xiaoqinli.git
cd xiaoqinli && go build -o xql .
# Move to system PATH (macOS/Linux example)
sudo mv xql /usr/local/bin/
Verifying Installation
Run the following command to check help and availability:
xql --help
Xiaoqinli (XQL) Data Schema & Taxonomy
Type System Mappings
All AST components in XQL resolve to underlying target types across languages:
| AST Kind | Go Type | Rust Type | TypeScript Type | Kotlin Type | Python Type |
|---|---|---|---|---|---|
Int |
int |
i64 |
number |
Long |
int |
Float |
float64 |
f64 |
number |
Double |
float |
String |
string |
String |
string |
String |
str |
Bool |
bool |
bool |
boolean |
Boolean |
bool |
Void |
(none) | (none) | void |
Unit |
None |
Array |
[]T |
Vec<T> |
T[] |
List<T> |
list[T] |
Option |
*T |
Option<T> |
T | null |
T? |
Optional[T] |
AST Node Taxonomy
Every .xql.json object contains a "kind" field. Declarations specify Program and FunctionDecl. Statement nodes structure variables and control loops (VarDecl, AssignStmt, IfStmt, WhileStmt). Expressions formulate variables and operations (Literal, Ident, BinaryExpr, CallExpr).
Xiaoqinli (XQL) Advanced Features
- Capability Sandboxing: Use target capability annotations (
grant) to restrict file system, external network, or standard IO access. - Static Effect Inference: Programmatically restrict function behaviors to strictly pure paths (
effects: ["pure"]) for mathematical models and calculations. - Multi-Language Transpilation Target: Emit verified schemas to six major industry languages natively from the exact same JSON configuration.
- Detailed Compiler Diagnostic Codes: Easily parse compile-time failures using distinct error codes for schema errors (
XQL_E1xx), type conflicts (XQL_E2xx), capability denials (XQL_E3xx), and output generation glitches (XQL_E4xx).
SKILL.md
Loading
Related Openclaw Skills
Viral Content Score
An AI-powered copywriting evaluation engine that predicts content virality across TikTok, Twitter, YouTube Shorts, and Instagram.
Cross-Cultural Insight
An AI-powered global brand compliance engine that detects linguistic, visual, and cultural risks across 50+ international markets.
Nefrologia Clinica (Clinical Nephrology)
An AI-powered clinical nephrology assistant supporting literature searches, decision-making, and clinical reasoning in kidney diseases, acid-base imbalances, and renal replacement therapies.
Global Market Trend Analysis
An AI-driven market intelligence skill that provides real-time track trends, quantitative growth metrics, and regional comparisons for strategic decision-making.
Blender Python Automation (bpy)
Automate 3D asset generation, procedural modeling, rendering, and 3D print preparation using Blender's Python API.
AI Startup Scout
A real-time investment and market intelligence scouting tool for tracking global AI startups, funding rounds, and founding team backgrounds.
Featured*
