purestrata 0.1.0__tar.gz

purestrata-0.1.0/.gitignore

@@ -0,0 +1,12 @@
+__pycache__/
+*.py[cod]
+.venv/
+dist/
+build/
+*.egg-info/
+.pytest_cache/
+.ruff_cache/
+examples/corpus/
+examples/output/
+.env
+.DS_Store

purestrata-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Daiki Ito
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

purestrata-0.1.0/PKG-INFO

@@ -0,0 +1,227 @@
+Metadata-Version: 2.5
+Name: purestrata
+Version: 0.1.0
+Summary: Open data layer for manufacturing AI: structure drawings, spec sheets, and failure knowledge into LLM tools and fine-tuning datasets.
+Project-URL: Homepage, https://github.com/daikiito-dk/PureStrata
+Project-URL: Repository, https://github.com/daikiito-dk/PureStrata
+Author: Daiki Ito
+License-Expression: MIT
+License-File: LICENSE
+Keywords: cad,dxf,failure-analysis,fine-tuning,llm,manuals,manufacturing,mcp,rag,tool-calling
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Developers
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Scientific/Engineering
+Classifier: Topic :: Text Processing
+Requires-Python: >=3.10
+Provides-Extra: mcp
+Requires-Dist: mcp>=1.2; extra == 'mcp'
+Provides-Extra: pdf
+Requires-Dist: pypdf>=4.0; extra == 'pdf'
+Description-Content-Type: text/markdown
+
+# PureStrata
+
+**English** | [日本語](#日本語)
+
+The open data layer for manufacturing AI. Ingest **drawings, spec sheets, and failure knowledge** into one portable, provenance-tagged corpus — query it from any LLM through MCP or tool calling, then export the same corpus as **fine-tuning datasets**.
+
+Manufacturing is where LLMs struggle most: CAD drawings, dense maintenance manuals, and failure knowledge locked in spreadsheets. Japanese factories in particular hold decades of primary data — drawings, 保全記録, and skilled-worker know-how — that never reaches the open web. PureStrata turns those documents into a structured, auditable layer that model developers and AI teams can actually use.
+
+```
+[ manuals (MD/PDF) ]   [ drawings (DXF) ]   [ failure logs (CSV/JSONL) ]
+        │                     │                      │
+        └─────────────────────┴──────────────────────┘
+                              ▼
+              [ ingest ] clean → structure → provenance
+                              ▼
+              [ corpus ]  portable *.strata.jsonl files
+                              ▼
+        ┌─────────────────────┴─────────────────────┐
+        ▼                       ▼                    ▼
+ [ MCP server ]      [ tool calling (Grok,    [ fine-tuning datasets ]
+ any MCP client        OpenAI, local LLMs) ]     chat / tool-trace / CoT
+```
+
+## Why this exists
+
+- **LLMs are weak on manufacturing data.** PDF drawings, CAD files, and complex spec sheets are exactly the data that never made it into pretraining corpora.
+- **Failure knowledge is the missing stratum.** Most RAG stacks retrieve manual text. Factory troubleshooting runs on *symptom → cause → countermeasure* records. PureStrata treats failure cases as first-class structured data.
+- **Provenance is non-negotiable.** Every record carries its source file and locator, so any answer an LLM gives can be audited back to the drawing, section, or spreadsheet row it came from.
+- **No lock-in.** The corpus is plain JSONL you own. Zero required dependencies for the core. Runs on your machine, air-gap friendly.
+
+## Install
+
+```bash
+uv add purestrata            # or: pip install purestrata
+pip install "purestrata[pdf]"    # optional: PDF manual support (pypdf)
+pip install "purestrata[mcp]"    # optional: MCP server support
+```
+
+Python 3.10+ required.
+
+## Quickstart
+
+```bash
+git clone https://github.com/daikiito-dk/PureStrata
+cd PureStrata
+make setup
+make demo          # ingest the bundled sample data, then search it
+```
+
+Or step by step:
+
+```bash
+# 1. Ingest manuals, a drawing, and a failure log into a corpus
+purestrata ingest-manual examples/data/manuals/lathe_mk3_maintenance.md --out corpus/
+purestrata ingest-manual examples/data/manuals/ja/cnc_press_maintenance.md --out corpus/
+purestrata ingest-dxf examples/data/cad/bracket_mk3.dxf --out corpus/
+purestrata ingest-failures examples/data/failures/plant_failures.csv --out corpus/
+
+# 2. Inspect and search (English and Japanese)
+purestrata inspect --corpus corpus/
+purestrata search "spindle vibration" --corpus corpus/
+purestrata search "ブレーキ 停止時間" --corpus corpus/
+
+# 3. Export fine-tuning datasets from the same corpus
+purestrata export chat       --corpus corpus/ --out dataset_chat.jsonl
+purestrata export tool_trace --corpus corpus/ --out dataset_tools.jsonl
+purestrata export cot        --corpus corpus/ --out dataset_cot.jsonl
+```
+
+## The three record families
+
+| Kind | Source | What the LLM gets |
+| --- | --- | --- |
+| `manual_section` | Markdown / PDF manuals | Clean, chunked spec-sheet passages with section paths |
+| `procedure` | Numbered lists in manuals | Ordered maintenance steps as structured arrays |
+| `safety_notice` | WARNING / CAUTION lines (EN + JA) | Severity-tagged precautions |
+| `part_record` | Parts tables | Part number, name, quantity, spec |
+| `drawing` | ASCII DXF files | Title block, layers, entity counts, annotations, bounds |
+| `failure_case` | CSV / JSONL failure logs | Symptom, root cause, countermeasure, severity, tags |
+
+## Tools (identical over MCP, OpenAI-style calls, and Python)
+
+| Tool | Purpose |
+| --- | --- |
+| `list_corpus` | Inventory: record counts by kind, loaded documents |
+| `search_documents` | Ranked passage search (English and Japanese) |
+| `get_safety_notices` | Warnings and cautions, optionally ranked |
+| `get_procedure` | Full ordered steps of one procedure |
+| `search_parts` | Parts lookup from parts lists |
+| `list_drawings` / `get_drawing` | CAD inventory and full drawing metadata |
+| `search_failure_cases` | **Troubleshooting first:** match symptoms against past failures |
+| `get_failure_case` | One full case by ID |
+
+```bash
+purestrata tools openai      # JSON schemas ready for Grok / OpenAI / compatible APIs
+purestrata tools anthropic
+```
+
+## Integrations
+
+**Grok / OpenAI / any OpenAI-compatible API** — pass the tool schemas to the chat completions endpoint and execute returned tool calls locally. See `examples/api_tool_use.py`:
+
+```bash
+export PURESTRATA_BASE_URL=https://api.x.ai/v1
+export PURESTRATA_API_KEY=your-key
+uv run python examples/api_tool_use.py "The lathe is vibrating during cutting. What should we check?"
+```
+
+Your corpus is queried locally; results go back to the model with provenance.
+
+**MCP (Model Context Protocol)** — run the bundled server and register it with any MCP client:
+
+```json
+{ "mcpServers": { "purestrata": { "command": "purestrata", "args": ["serve", "--corpus", "/path/to/corpus"] } } }
+```
+
+**Python API** — the same tools as a library:
+
+```python
+import purestrata as ps
+
+chunks = ps.ingest_manual("manual.md") + ps.ingest_failures("failures.csv")
+ps.write_chunks(chunks, "corpus/", "my_doc")
+executor = ps.ToolExecutor(ps.Corpus.load("corpus/"))
+hits = executor.execute("search_failure_cases", {"query": "spindle vibration", "top_k": 3})
+```
+
+## Fine-tuning exports
+
+The same corpus that serves live queries becomes training data:
+
+- **chat** — system/user/assistant JSONL accepted by OpenAI's and xAI's fine-tuning APIs
+- **tool_trace** — model-issued tool call → provenance-tagged tool result → grounded answer
+- **cot** — instruction / reasoning / output records for chain-of-thought training
+
+CoT records are a deliberate scaffold: they are transparent, grounded in structured fields, and built to be **reviewed and sharpened by maintenance engineers** before training. That expert-in-the-loop pass is where dataset quality comes from — PureStrata keeps the work auditable so it can be done rigorously.
+
+## Status and roadmap
+
+Current: local ingestion (Markdown, PDF, ASCII DXF, CSV/JSONL), BM25 search with Japanese support, MCP server, tool schemas, fine-tune exports.
+
+Planned: HTML/DXF-R2018 support, STEP files, per-layer geometry search, embedding search, evaluation harness for fine-tune sets, and a hosted SaaS for teams that need managed ingestion, access control, and expert review workflows.
+
+## Development
+
+```bash
+make setup    # uv sync (installs dev tools and all extras)
+make lint     # ruff check + format check
+make test     # pytest
+```
+
+## License
+
+[MIT](LICENSE)
+
+---
+
+## 日本語
+
+PureStrata は製造業AIのためのオープンなデータレイヤーです。**図面・仕様書・不具合ナレッジ**を1つのポータブルで出所証明（プロビナンス）付きのコーパスに取り込み、MCPやツール呼び出しでどのLLMからでも検索でき、同じコーパスから**微調整用データセット**を書き出せます。
+
+製造業はLLMが最も苦手とする領域です。CAD図面、高密度な保全マニュアル、そしてスプレッドシートに閉じ込められた不具合ナレッジ。特に日本の工場には、図面・保全記録・熟練工のノウハウという、Web上に存在しない一次データが蓄積されています。PureStrataはそれらの文書を、構造化され監査可能なデータ層に変換します。
+
+### 特徴
+
+- **3つのレコード族を第一級データとして扱う**: 仕様書（`manual_section` / `procedure` / `safety_notice` / `part_record`）、図面（DXFタイトルブロック・レイヤー・注記）、不具合ナレッジ（現象→原因→対策）。特に「不具合ナレッジ」は通常のRAGには欠けている層で、工場のトラブルシューティングの中核です。
+- **完全な出所証明**: すべてのレコードが元ファイルと位置情報を保持し、LLMの回答がどの図面・セクション・行に遡れるか監査できます。
+- **ロックインなし**: コーパスはプレーンなJSONL。コアは依存ゼロ。ローカル実行でエアギップ環境にも対応します。
+- **日本語対応検索**: 依存ゼロのBM25実装がCJKバイグラム解析に対応し、日本語の不具合記録もそのまま検索できます。
+- **Grok / OpenAI / ローカルLLM対応**: OpenAI互換のツール呼び出し用スキーマと、MCPサーバーの両方を同梱。特定モデルへの依存はありません。
+
+### クイックスタート
+
+```bash
+git clone https://github.com/daikiito-dk/PureStrata
+cd PureStrata
+make setup
+make demo          # サンプルデータを取り込んで検索
+
+# 手順で実行する場合
+purestrata ingest-manual examples/data/manuals/ja/cnc_press_maintenance.md --out corpus/
+purestrata ingest-dxf examples/data/cad/bracket_mk3.dxf --out corpus/
+purestrata ingest-failures examples/data/failures/plant_failures.csv --out corpus/
+purestrata search "ブレーキ 停止時間" --corpus corpus/
+purestrata export cot --corpus corpus/ --out dataset_cot.jsonl
+```
+
+### 微調整用データセット
+
+`chat`（OpenAI / xAIファインチューニングAPI対応のメッセージ形式）、`tool_trace`（ツール呼び出し→出所付き結果→根拠ある回答）、`cot`（推論過程付き問答）の3形式で書き出せます。CoTレコードは構造化フィールドから生成される透明な雛形であり、保全エンジニアによるレビュー・推敲（ヒューマン・イン・ザ・ループ）を前提に設計されています。データ品質はその専門家レビューで決まるため、PureStrataはその作業を監査可能な形で支えます。
+
+### 開発
+
+```bash
+make setup    # uv sync
+make lint     # ruff
+make test     # pytest
+```
+
+### ライセンス
+
+[MIT](LICENSE)

purestrata-0.1.0/README.md

@@ -0,0 +1,202 @@
+# PureStrata
+
+**English** | [日本語](#日本語)
+
+The open data layer for manufacturing AI. Ingest **drawings, spec sheets, and failure knowledge** into one portable, provenance-tagged corpus — query it from any LLM through MCP or tool calling, then export the same corpus as **fine-tuning datasets**.
+
+Manufacturing is where LLMs struggle most: CAD drawings, dense maintenance manuals, and failure knowledge locked in spreadsheets. Japanese factories in particular hold decades of primary data — drawings, 保全記録, and skilled-worker know-how — that never reaches the open web. PureStrata turns those documents into a structured, auditable layer that model developers and AI teams can actually use.
+
+```
+[ manuals (MD/PDF) ]   [ drawings (DXF) ]   [ failure logs (CSV/JSONL) ]
+        │                     │                      │
+        └─────────────────────┴──────────────────────┘
+                              ▼
+              [ ingest ] clean → structure → provenance
+                              ▼
+              [ corpus ]  portable *.strata.jsonl files
+                              ▼
+        ┌─────────────────────┴─────────────────────┐
+        ▼                       ▼                    ▼
+ [ MCP server ]      [ tool calling (Grok,    [ fine-tuning datasets ]
+ any MCP client        OpenAI, local LLMs) ]     chat / tool-trace / CoT
+```
+
+## Why this exists
+
+- **LLMs are weak on manufacturing data.** PDF drawings, CAD files, and complex spec sheets are exactly the data that never made it into pretraining corpora.
+- **Failure knowledge is the missing stratum.** Most RAG stacks retrieve manual text. Factory troubleshooting runs on *symptom → cause → countermeasure* records. PureStrata treats failure cases as first-class structured data.
+- **Provenance is non-negotiable.** Every record carries its source file and locator, so any answer an LLM gives can be audited back to the drawing, section, or spreadsheet row it came from.
+- **No lock-in.** The corpus is plain JSONL you own. Zero required dependencies for the core. Runs on your machine, air-gap friendly.
+
+## Install
+
+```bash
+uv add purestrata            # or: pip install purestrata
+pip install "purestrata[pdf]"    # optional: PDF manual support (pypdf)
+pip install "purestrata[mcp]"    # optional: MCP server support
+```
+
+Python 3.10+ required.
+
+## Quickstart
+
+```bash
+git clone https://github.com/daikiito-dk/PureStrata
+cd PureStrata
+make setup
+make demo          # ingest the bundled sample data, then search it
+```
+
+Or step by step:
+
+```bash
+# 1. Ingest manuals, a drawing, and a failure log into a corpus
+purestrata ingest-manual examples/data/manuals/lathe_mk3_maintenance.md --out corpus/
+purestrata ingest-manual examples/data/manuals/ja/cnc_press_maintenance.md --out corpus/
+purestrata ingest-dxf examples/data/cad/bracket_mk3.dxf --out corpus/
+purestrata ingest-failures examples/data/failures/plant_failures.csv --out corpus/
+
+# 2. Inspect and search (English and Japanese)
+purestrata inspect --corpus corpus/
+purestrata search "spindle vibration" --corpus corpus/
+purestrata search "ブレーキ 停止時間" --corpus corpus/
+
+# 3. Export fine-tuning datasets from the same corpus
+purestrata export chat       --corpus corpus/ --out dataset_chat.jsonl
+purestrata export tool_trace --corpus corpus/ --out dataset_tools.jsonl
+purestrata export cot        --corpus corpus/ --out dataset_cot.jsonl
+```
+
+## The three record families
+
+| Kind | Source | What the LLM gets |
+| --- | --- | --- |
+| `manual_section` | Markdown / PDF manuals | Clean, chunked spec-sheet passages with section paths |
+| `procedure` | Numbered lists in manuals | Ordered maintenance steps as structured arrays |
+| `safety_notice` | WARNING / CAUTION lines (EN + JA) | Severity-tagged precautions |
+| `part_record` | Parts tables | Part number, name, quantity, spec |
+| `drawing` | ASCII DXF files | Title block, layers, entity counts, annotations, bounds |
+| `failure_case` | CSV / JSONL failure logs | Symptom, root cause, countermeasure, severity, tags |
+
+## Tools (identical over MCP, OpenAI-style calls, and Python)
+
+| Tool | Purpose |
+| --- | --- |
+| `list_corpus` | Inventory: record counts by kind, loaded documents |
+| `search_documents` | Ranked passage search (English and Japanese) |
+| `get_safety_notices` | Warnings and cautions, optionally ranked |
+| `get_procedure` | Full ordered steps of one procedure |
+| `search_parts` | Parts lookup from parts lists |
+| `list_drawings` / `get_drawing` | CAD inventory and full drawing metadata |
+| `search_failure_cases` | **Troubleshooting first:** match symptoms against past failures |
+| `get_failure_case` | One full case by ID |
+
+```bash
+purestrata tools openai      # JSON schemas ready for Grok / OpenAI / compatible APIs
+purestrata tools anthropic
+```
+
+## Integrations
+
+**Grok / OpenAI / any OpenAI-compatible API** — pass the tool schemas to the chat completions endpoint and execute returned tool calls locally. See `examples/api_tool_use.py`:
+
+```bash
+export PURESTRATA_BASE_URL=https://api.x.ai/v1
+export PURESTRATA_API_KEY=your-key
+uv run python examples/api_tool_use.py "The lathe is vibrating during cutting. What should we check?"
+```
+
+Your corpus is queried locally; results go back to the model with provenance.
+
+**MCP (Model Context Protocol)** — run the bundled server and register it with any MCP client:
+
+```json
+{ "mcpServers": { "purestrata": { "command": "purestrata", "args": ["serve", "--corpus", "/path/to/corpus"] } } }
+```
+
+**Python API** — the same tools as a library:
+
+```python
+import purestrata as ps
+
+chunks = ps.ingest_manual("manual.md") + ps.ingest_failures("failures.csv")
+ps.write_chunks(chunks, "corpus/", "my_doc")
+executor = ps.ToolExecutor(ps.Corpus.load("corpus/"))
+hits = executor.execute("search_failure_cases", {"query": "spindle vibration", "top_k": 3})
+```
+
+## Fine-tuning exports
+
+The same corpus that serves live queries becomes training data:
+
+- **chat** — system/user/assistant JSONL accepted by OpenAI's and xAI's fine-tuning APIs
+- **tool_trace** — model-issued tool call → provenance-tagged tool result → grounded answer
+- **cot** — instruction / reasoning / output records for chain-of-thought training
+
+CoT records are a deliberate scaffold: they are transparent, grounded in structured fields, and built to be **reviewed and sharpened by maintenance engineers** before training. That expert-in-the-loop pass is where dataset quality comes from — PureStrata keeps the work auditable so it can be done rigorously.
+
+## Status and roadmap
+
+Current: local ingestion (Markdown, PDF, ASCII DXF, CSV/JSONL), BM25 search with Japanese support, MCP server, tool schemas, fine-tune exports.
+
+Planned: HTML/DXF-R2018 support, STEP files, per-layer geometry search, embedding search, evaluation harness for fine-tune sets, and a hosted SaaS for teams that need managed ingestion, access control, and expert review workflows.
+
+## Development
+
+```bash
+make setup    # uv sync (installs dev tools and all extras)
+make lint     # ruff check + format check
+make test     # pytest
+```
+
+## License
+
+[MIT](LICENSE)
+
+---
+
+## 日本語
+
+PureStrata は製造業AIのためのオープンなデータレイヤーです。**図面・仕様書・不具合ナレッジ**を1つのポータブルで出所証明（プロビナンス）付きのコーパスに取り込み、MCPやツール呼び出しでどのLLMからでも検索でき、同じコーパスから**微調整用データセット**を書き出せます。
+
+製造業はLLMが最も苦手とする領域です。CAD図面、高密度な保全マニュアル、そしてスプレッドシートに閉じ込められた不具合ナレッジ。特に日本の工場には、図面・保全記録・熟練工のノウハウという、Web上に存在しない一次データが蓄積されています。PureStrataはそれらの文書を、構造化され監査可能なデータ層に変換します。
+
+### 特徴
+
+- **3つのレコード族を第一級データとして扱う**: 仕様書（`manual_section` / `procedure` / `safety_notice` / `part_record`）、図面（DXFタイトルブロック・レイヤー・注記）、不具合ナレッジ（現象→原因→対策）。特に「不具合ナレッジ」は通常のRAGには欠けている層で、工場のトラブルシューティングの中核です。
+- **完全な出所証明**: すべてのレコードが元ファイルと位置情報を保持し、LLMの回答がどの図面・セクション・行に遡れるか監査できます。
+- **ロックインなし**: コーパスはプレーンなJSONL。コアは依存ゼロ。ローカル実行でエアギップ環境にも対応します。
+- **日本語対応検索**: 依存ゼロのBM25実装がCJKバイグラム解析に対応し、日本語の不具合記録もそのまま検索できます。
+- **Grok / OpenAI / ローカルLLM対応**: OpenAI互換のツール呼び出し用スキーマと、MCPサーバーの両方を同梱。特定モデルへの依存はありません。
+
+### クイックスタート
+
+```bash
+git clone https://github.com/daikiito-dk/PureStrata
+cd PureStrata
+make setup
+make demo          # サンプルデータを取り込んで検索
+
+# 手順で実行する場合
+purestrata ingest-manual examples/data/manuals/ja/cnc_press_maintenance.md --out corpus/
+purestrata ingest-dxf examples/data/cad/bracket_mk3.dxf --out corpus/
+purestrata ingest-failures examples/data/failures/plant_failures.csv --out corpus/
+purestrata search "ブレーキ 停止時間" --corpus corpus/
+purestrata export cot --corpus corpus/ --out dataset_cot.jsonl
+```
+
+### 微調整用データセット
+
+`chat`（OpenAI / xAIファインチューニングAPI対応のメッセージ形式）、`tool_trace`（ツール呼び出し→出所付き結果→根拠ある回答）、`cot`（推論過程付き問答）の3形式で書き出せます。CoTレコードは構造化フィールドから生成される透明な雛形であり、保全エンジニアによるレビュー・推敲（ヒューマン・イン・ザ・ループ）を前提に設計されています。データ品質はその専門家レビューで決まるため、PureStrataはその作業を監査可能な形で支えます。
+
+### 開発
+
+```bash
+make setup    # uv sync
+make lint     # ruff
+make test     # pytest
+```
+
+### ライセンス
+
+[MIT](LICENSE)

purestrata-0.1.0/examples/api_tool_use.py

@@ -0,0 +1,85 @@
+"""Query a PureStrata corpus from Grok (or any OpenAI-compatible API) with tool calling.
+
+The PureStrata tool schemas are passed to the chat completions endpoint;
+tool calls are executed locally against your corpus, and results are sent
+back with provenance. Your data never leaves your machine except through
+the API you configure.
+
+    export PURESTRATA_BASE_URL=https://api.x.ai/v1      # or https://api.openai.com/v1
+    export PURESTRATA_API_KEY=your-key
+    export PURESTRATA_MODEL=grok-4                     # model id of your provider
+    uv run python examples/api_tool_use.py "The lathe is vibrating during cutting. What should we check?"
+"""
+
+from __future__ import annotations
+
+import json
+import os
+import sys
+import urllib.request
+from pathlib import Path
+
+from purestrata import Corpus, ToolError, ToolExecutor, openai_tools
+
+HERE = Path(__file__).resolve().parent
+
+
+def call_llm(base_url: str, api_key: str, model: str, messages: list[dict], tools: list[dict]) -> dict:
+    payload = {"model": model, "messages": messages, "tools": tools}
+    request = urllib.request.Request(
+        f"{base_url.rstrip('/')}/chat/completions",
+        data=json.dumps(payload).encode("utf-8"),
+        headers={"Content-Type": "application/json", "Authorization": f"Bearer {api_key}"},
+    )
+    with urllib.request.urlopen(request, timeout=120) as response:
+        return json.loads(response.read())
+
+
+def main() -> None:
+    base_url = os.environ.get("PURESTRATA_BASE_URL", "https://api.x.ai/v1")
+    api_key = os.environ.get("PURESTRATA_API_KEY", "")
+    model = os.environ.get("PURESTRATA_MODEL", "grok-4")
+    question = (
+        sys.argv[1]
+        if len(sys.argv) > 1
+        else "The lathe is vibrating during cutting. What should we check?"
+    )
+    corpus_dir = os.environ.get("PURESTRATA_CORPUS", str(HERE / "corpus"))
+    if not api_key:
+        print("Set PURESTRATA_API_KEY to run this example.", file=sys.stderr)
+        sys.exit(1)
+
+    executor = ToolExecutor(Corpus.load(corpus_dir))
+    tools = openai_tools()
+    messages: list[dict] = [{"role": "user", "content": question}]
+
+    for _round in range(5):
+        data = call_llm(base_url, api_key, model, messages, tools)
+        message = data["choices"][0]["message"]
+        messages.append(message)
+        tool_calls = message.get("tool_calls") or []
+        if not tool_calls:
+            print(message.get("content") or "(no content returned)")
+            return
+        for call in tool_calls:
+            name = call["function"]["name"]
+            try:
+                args = json.loads(call["function"]["arguments"] or "{}")
+            except json.JSONDecodeError:
+                args = {}
+            print(f"-> {name}({args})")
+            try:
+                result = executor.execute(name, args)
+            except ToolError as exc:
+                result = {"error": str(exc)}
+            messages.append(
+                {
+                    "role": "tool",
+                    "tool_call_id": call["id"],
+                    "content": json.dumps(result, ensure_ascii=False, default=str),
+                }
+            )
+
+
+if __name__ == "__main__":
+    main()