S
shahriar-ahmed-seam/LexCompiler
Neuro-Symbolic Law-to-AST Compiler, Microsoft Z3 Formal Verifier, 12-MCP Suite, Contextual+Graph RAG, and…
在 GitHub 查看 ↗链接已于 2026-11-12 核查 · HTTP 200 · 打开的是厂商自己的官网
安装与配置
客户端配置（取自该仓库的安装命令）
{
"mcpServers": {
"shahriar-ahmed-seam-lexcompiler": {
"command": "npx",
"args": ["tsx"]
}
}
}解析自我们于 2026-10-28 抓取的该仓库 README；我们未执行过。
command/args 来自仓库写明的安装命令；不确定时请以该项目 README 为准。
VS Code（.vscode/mcp.json —— 键名是 servers）
{
"servers": {
"shahriar-ahmed-seam-lexcompiler": {
"command": "npx",
"args": ["tsx"]
}
}
}VS Code 顶层键是 <code>servers</code>，其余与上面那份相同。
Claude Code（命令行）
claude mcp add --transport stdio shahriar-ahmed-seam-lexcompiler -- npx tsx
Claude Code 不收 JSON 文件，用这条命令直接加。
配置项的键名以你的客户端为准：客户端与 CLI
- ★ 数取值
- 3 ★ · shahriar-ahmed-seam/LexCompiler · 取自 GitHub 搜索 API · （2026-11-13 20:00:50 核对）
- 来源明细
- 从该仓库的 GitHub 页面直接读取 · 条目类型：开源项目
- 收录与活跃
- 收录于 2026-10-28 · 最近见到 2026-11-14 · 共见到 19 次 · 被 1 个来源列出
- 评分构成
- 综合分 103 · 3 ★ · 1 个来源 · repo
- 安装
npx tsx
常见问题（答案都来自我们的记录）
怎么安装？
该仓库给出的安装命令是
npx tsx。我们的记录里它是仓库自己提供的；不保证在你的客户端上直接可用。它适合谁 / 用来做什么？
我们把它归在这些标签下: 开源项目。标签来自条目自身，不是我们强加的分类。
谁维护它、还在更新吗？
仓库
shahriar-ahmed-seam/LexCompiler；我们收录于 2026-10-28；最近见到 2026-11-14。这里没有"官方认证"或"质量评级"——我们只记录来源与时间。# LexCompiler
以下摘录自该仓库的 README（片段） · 在 GitHub 读完整 README →
Neuro-Symbolic Law-to-AST Compiler, SMT/Z3 Formal Verifier, and Autonomous Pull Request Enforcer
Domain: JusticeSphere AI / ComplianceOS / GovernanceIQ | Engine: Microsoft Z3 v4.13.4 + OPA Rego + Polyglot AST | Protocol: MCP JSON-RPC 2.0 (2025-03-26)
## Abstract
Enterprise software compliance suffers from a structural impedance mismatch: regulatory statutes and enterprise Data Processing Agreements (DPAs) are authored in natural language, whereas production systems execute as deterministic code and relational schemas. Legal teams interpret regulatory updates in static documents while engineering teams ship hundreds of commits per week across API routes, caching workers, and database schemas.
LexCompiler eliminates the manual translation layer between statutory text and source code. It ingests live regulatory feeds (GDPR, HIPAA, EU AI Act, CCPA/CPRA, EUR-Lex, U.S. Federal Register) and custom enterprise contracts, compiles normative clauses into modal deontic logic predicates, synthesizes machine-verifiable SMT-LIB2 assertions and Open Policy Agent (OPA / Rego) rules, and evaluates them against a repository's live Abstract Syntax Tree (AST) and SQL DDL schemas.
When a statutory clause is amended or a code commit introduces a compliance regression, LexCompiler:
1. Proves the mathematical contradiction (UNSAT) using the Microsoft Z3 Theorem Prover and extracts the minimal Unsat Core and Counterexample Witness Model.
2. Pinpoints the exact offending AST nodes and line ranges across SQL, Python, and TypeScript modules.
3. Synthesizes an AST-preserving code patch via Counterexample-Guided Inductive Synthesis (CEGIS), verifies in an isolated Z3 sandbox that the patched AST transitions from UNSAT to SAT, and commits the remediation branch with an RFC 8032 Ed25519 cryptographic conformity certificate and OASIS SARIF v2.1.0 report.
## System Architecture
mermaid
flowchart TB
subgraph L1["1. Normative & Statutory Ingestion Layer"]
EUR["EUR-Lex SPARQL / REST Feed"]
FED["U.S. Federal Register API"]
DPA["Enterprise DPA & MSA Clauses"]
BYOC["Universal BYOC Custom Prose"]
end
subgraph L2["2. Deontic & Neuro-Symbolic Compilation Layer (lex-deontic-compiler-mcp)"]
DEONTIC["Modal Deontic Extractor: O(phi), F(phi), P(phi)"]
FOL["First-Order Predicate Logic Normalizer"]
SMT["SMT-LIB2 Generator (QFLIA & QFUF Named Assertions)"]
REGO["Open Policy Agent (OPA Rego) deny[msg] Compiler"]
ASTQ["Polyglot AST & SQL DDL S-Expression Query Generator"]
end
subgraph L3["3. Polyglot AST & Runtime Introspection Layer (lex-polyglot-ast-mcp)"]
SQL["PostgreSQL / SQLite DDL Parser (schema/postgrespii.sql)"]
PY["Python 3.12 Native ast Parser (api/fastapiintake.py, workers/rediscacheworker.py)"]
TS["TypeScript 5.7 Compiler API (services/aicredittriage.ts, services/crossbordersync.ts)"]
SANDBOX["Isolated Runtime Sandbox (SQLite3 + Python 3.12 + Git Worktree)"]
end
subgraph L4["4. Formal Verification, Multi-LLM & Hybrid RAG Mesh"]
Z3["Microsoft Z3 Theorem Prover v4.13.4 (check-sat, get-unsat-core, get-model)"]
RAG["Contextual + Graph RAG (all-minilm 384-d Neural Embeddings + BM25 RRF + Neo4j Cypher)"]
LLM["Local Ollama Multi-LLM Daemon (:11434) (qwen2.5-coder:0.5b + dolphin3:latest 8B)"]
Domain: JusticeSphere AI / ComplianceOS / GovernanceIQ | Engine: Microsoft Z3 v4.13.4 + OPA Rego + Polyglot AST | Protocol: MCP JSON-RPC 2.0 (2025-03-26)
## Abstract
Enterprise software compliance suffers from a structural impedance mismatch: regulatory statutes and enterprise Data Processing Agreements (DPAs) are authored in natural language, whereas production systems execute as deterministic code and relational schemas. Legal teams interpret regulatory updates in static documents while engineering teams ship hundreds of commits per week across API routes, caching workers, and database schemas.
LexCompiler eliminates the manual translation layer between statutory text and source code. It ingests live regulatory feeds (GDPR, HIPAA, EU AI Act, CCPA/CPRA, EUR-Lex, U.S. Federal Register) and custom enterprise contracts, compiles normative clauses into modal deontic logic predicates, synthesizes machine-verifiable SMT-LIB2 assertions and Open Policy Agent (OPA / Rego) rules, and evaluates them against a repository's live Abstract Syntax Tree (AST) and SQL DDL schemas.
When a statutory clause is amended or a code commit introduces a compliance regression, LexCompiler:
1. Proves the mathematical contradiction (UNSAT) using the Microsoft Z3 Theorem Prover and extracts the minimal Unsat Core and Counterexample Witness Model.
2. Pinpoints the exact offending AST nodes and line ranges across SQL, Python, and TypeScript modules.
3. Synthesizes an AST-preserving code patch via Counterexample-Guided Inductive Synthesis (CEGIS), verifies in an isolated Z3 sandbox that the patched AST transitions from UNSAT to SAT, and commits the remediation branch with an RFC 8032 Ed25519 cryptographic conformity certificate and OASIS SARIF v2.1.0 report.
## System Architecture
mermaid
flowchart TB
subgraph L1["1. Normative & Statutory Ingestion Layer"]
EUR["EUR-Lex SPARQL / REST Feed"]
FED["U.S. Federal Register API"]
DPA["Enterprise DPA & MSA Clauses"]
BYOC["Universal BYOC Custom Prose"]
end
subgraph L2["2. Deontic & Neuro-Symbolic Compilation Layer (lex-deontic-compiler-mcp)"]
DEONTIC["Modal Deontic Extractor: O(phi), F(phi), P(phi)"]
FOL["First-Order Predicate Logic Normalizer"]
SMT["SMT-LIB2 Generator (QFLIA & QFUF Named Assertions)"]
REGO["Open Policy Agent (OPA Rego) deny[msg] Compiler"]
ASTQ["Polyglot AST & SQL DDL S-Expression Query Generator"]
end
subgraph L3["3. Polyglot AST & Runtime Introspection Layer (lex-polyglot-ast-mcp)"]
SQL["PostgreSQL / SQLite DDL Parser (schema/postgrespii.sql)"]
PY["Python 3.12 Native ast Parser (api/fastapiintake.py, workers/rediscacheworker.py)"]
TS["TypeScript 5.7 Compiler API (services/aicredittriage.ts, services/crossbordersync.ts)"]
SANDBOX["Isolated Runtime Sandbox (SQLite3 + Python 3.12 + Git Worktree)"]
end
subgraph L4["4. Formal Verification, Multi-LLM & Hybrid RAG Mesh"]
Z3["Microsoft Z3 Theorem Prover v4.13.4 (check-sat, get-unsat-core, get-model)"]
RAG["Contextual + Graph RAG (all-minilm 384-d Neural Embeddings + BM25 RRF + Neo4j Cypher)"]
LLM["Local Ollama Multi-LLM Daemon (:11434) (qwen2.5-coder:0.5b + dolphin3:latest 8B)"]
收录徽章（可选）
想把收录标记放进 README 就复制这行，可选，不影响收录。
[](https://anclum.com/zh/s/shahriar-ahmed-seam-lexcompiler/)
- 来源
- github · https://github.com/shahriar-ahmed-seam/LexCompiler
- 按领域浏览
- 开源项目
- Traction
- 3 ★
- 收录于
- 今天
同类推荐
所属领域与标签
每个标签都能点进去看同类的全部条目。数字是当前库里的真实条数。