mailwise 0.1.0__tar.gz

mailwise-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Petr
+
+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.

mailwise-0.1.0/PKG-INFO

@@ -0,0 +1,206 @@
+Metadata-Version: 2.4
+Name: mailwise
+Version: 0.1.0
+Summary: Turn email threads into a searchable knowledge base. Parse EML files, index with embeddings, and use RAG to learn how your best engineers analyze issues.
+Author: Petr
+License-Expression: MIT
+Project-URL: Homepage, https://github.com/PetrGuan/MailWise
+Project-URL: Issues, https://github.com/PetrGuan/MailWise/issues
+Keywords: email,eml,rag,embeddings,issue-tracking,knowledge-base
+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 :: Communications :: Email
+Classifier: Topic :: Text Processing :: Indexing
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: sentence-transformers>=2.2.0
+Requires-Dist: numpy>=1.24.0
+Requires-Dist: click>=8.0
+Requires-Dist: pyyaml>=6.0
+Provides-Extra: dev
+Requires-Dist: pytest>=7.0; extra == "dev"
+Requires-Dist: pytest-cov>=4.0; extra == "dev"
+Dynamic: license-file
+
+# MailWise
+
+[![CI](https://github.com/PetrGuan/MailWise/actions/workflows/ci.yml/badge.svg)](https://github.com/PetrGuan/MailWise/actions/workflows/ci.yml)
+[![PyPI version](https://badge.fury.io/py/mailwise.svg)](https://pypi.org/project/mailwise/)
+[![Python 3.10+](https://img.shields.io/badge/python-3.10%2B-blue.svg)](https://python.org)
+[![License: MIT](https://img.shields.io/badge/License-MIT-green.svg)](LICENSE)
+
+[English](README.md) | [中文](README.zh-CN.md)
+
+Turn email threads into a searchable knowledge base. Parse EML files, index with embeddings, and use RAG to learn how your best engineers analyze issues.
+
+## What it does
+
+MailWise reads `.eml` files (exported from Outlook, Thunderbird, etc.), splits email threads into individual replies, and builds a semantic search index. You can then:
+
+- **Search** for similar past issues using natural language
+- **Analyze** new issues with RAG — Claude reads how your experts solved similar problems and synthesizes advice
+- **Tag expert engineers** whose replies get boosted in search results and highlighted in output
+
+## Why
+
+If your team handles bugs/incidents via email, years of tribal knowledge is buried in threads. MailWise makes that knowledge searchable and actionable.
+
+## Quick start
+
+### Prerequisites
+
+- Python 3.10+
+- [Claude Code](https://claude.ai/code) (for the `analyze` command — uses your existing auth, no API key needed)
+
+### Install
+
+From PyPI:
+
+```bash
+pip install mailwise
+```
+
+Or from source:
+
+```bash
+git clone https://github.com/PetrGuan/MailWise.git
+cd MailWise
+pip install -e .
+```
+
+### Configure
+
+The easiest way to get started:
+
+```bash
+mailwise init
+```
+
+This will walk you through setting up your EML directory, adding expert engineers, and verifying the setup.
+
+Or configure manually:
+
+```bash
+cp config.example.yaml config.yaml
+```
+
+Edit `config.yaml` with your settings:
+
+```yaml
+eml_directory: /path/to/your/eml/files
+database: data/index.db
+markdown_directory: markdown
+embedding_model: all-MiniLM-L6-v2
+expert_boost: 1.5
+
+experts:
+  - email: senior.dev@company.com
+    name: Jane Doe
+```
+
+### Usage
+
+```bash
+# Index your emails (incremental — only processes new/changed files)
+mailwise index
+
+# Search for similar past issues
+mailwise search "sync failure after folder migration"
+
+# Search with previews
+mailwise search "calendar not updating" --show-body
+
+# Only show expert replies
+mailwise search "deleted emails reappear" --expert-only
+
+# Deep analysis — Claude reasons over similar expert threads
+mailwise analyze "User reports emails moved to local folder keep reappearing in Inbox"
+
+# View full markdown of a specific email thread
+mailwise show 42
+
+# Check index stats
+mailwise stats
+```
+
+### Managing experts
+
+```bash
+# Add an expert
+mailwise experts add engineer@company.com --name "Jane Doe"
+
+# List all experts
+mailwise experts list
+
+# Remove an expert
+mailwise experts remove engineer@company.com
+```
+
+## How it works
+
+```
+EML files → Parser → Markdown + Embeddings → SQLite index
+                                                    ↓
+                              Query → Semantic search → Top matches
+                                                            ↓
+                                          Claude (via RAG) → Expert-informed analysis
+```
+
+1. **Parse**: EML files are parsed in parallel and threads are split into individual replies using Outlook-style `From:/Sent:` delimiters
+2. **Clean**: Microsoft SafeLinks are unwrapped, mailto artifacts are removed
+3. **Markdown**: Each thread becomes a structured markdown file with `[Expert]` tags on replies from your designated engineers
+4. **Embed**: Each reply is embedded using `all-MiniLM-L6-v2` (runs locally, no API calls)
+5. **Index**: Embeddings and metadata are stored in SQLite for fast retrieval
+6. **Search**: Cosine similarity with expert score boosting finds relevant past issues
+7. **Analyze**: Top matches are fed to Claude (via Claude Code CLI) with a system prompt that focuses on expert reasoning patterns
+
+## Performance
+
+Designed for large mailboxes (25K+ emails, 16GB+):
+
+| Operation | Performance |
+|---|---|
+| Incremental check (no changes) | ~2-3s for 25K files (stat-based, no file reads) |
+| Full index | ~5-10 min (parallel parsing + batch embedding) |
+| Search query | <100ms (single matrix multiply over 100K+ vectors) |
+| RAG analysis | ~10-20s (retrieval + Claude response) |
+
+Key optimizations:
+- **Two-phase change detection**: mtime+size stat check before SHA256 hashing
+- **Parallel EML parsing**: multiprocessing with configurable workers
+- **Batch embedding**: pre-computed offset arrays, no O(n²) lookups
+- **Optimized search**: loads only embedding BLOBs into contiguous numpy array; fetches metadata only for top-k results
+- **SQLite tuning**: WAL journal, 64MB cache, 256MB mmap, batch inserts via `executemany`
+
+## Architecture
+
+```
+src/email_issue_indexer/
+├── cli.py          # Click-based CLI
+├── parser.py       # EML parsing + thread splitting (parallel-safe)
+├── markdown.py     # Markdown conversion with expert tags
+├── safelinks.py    # Microsoft SafeLinks URL cleaning
+├── embeddings.py   # sentence-transformers embeddings + vector search
+├── store.py        # SQLite storage layer (performance-tuned)
+├── indexer.py      # Parallel batch orchestrator with progress tracking
+├── search.py       # Optimized similarity search with expert boosting
+└── rag.py          # RAG layer using Claude Code CLI
+```
+
+## Privacy
+
+All processing is local:
+- Embeddings run on your machine (no data sent to any API for indexing)
+- Email content stays in your local SQLite database and markdown files
+- The `analyze` command sends relevant thread excerpts to Claude — same as chatting in Claude Code
+
+Your `config.yaml`, `emails/`, `data/`, and `markdown/` directories are gitignored by default. Only `config.example.yaml` (with no real data) is committed. A pre-commit hook (`scripts/install-hooks.sh`) scans for accidental PII leaks.
+
+## License
+
+MIT

mailwise-0.1.0/README.md

@@ -0,0 +1,177 @@
+# MailWise
+
+[![CI](https://github.com/PetrGuan/MailWise/actions/workflows/ci.yml/badge.svg)](https://github.com/PetrGuan/MailWise/actions/workflows/ci.yml)
+[![PyPI version](https://badge.fury.io/py/mailwise.svg)](https://pypi.org/project/mailwise/)
+[![Python 3.10+](https://img.shields.io/badge/python-3.10%2B-blue.svg)](https://python.org)
+[![License: MIT](https://img.shields.io/badge/License-MIT-green.svg)](LICENSE)
+
+[English](README.md) | [中文](README.zh-CN.md)
+
+Turn email threads into a searchable knowledge base. Parse EML files, index with embeddings, and use RAG to learn how your best engineers analyze issues.
+
+## What it does
+
+MailWise reads `.eml` files (exported from Outlook, Thunderbird, etc.), splits email threads into individual replies, and builds a semantic search index. You can then:
+
+- **Search** for similar past issues using natural language
+- **Analyze** new issues with RAG — Claude reads how your experts solved similar problems and synthesizes advice
+- **Tag expert engineers** whose replies get boosted in search results and highlighted in output
+
+## Why
+
+If your team handles bugs/incidents via email, years of tribal knowledge is buried in threads. MailWise makes that knowledge searchable and actionable.
+
+## Quick start
+
+### Prerequisites
+
+- Python 3.10+
+- [Claude Code](https://claude.ai/code) (for the `analyze` command — uses your existing auth, no API key needed)
+
+### Install
+
+From PyPI:
+
+```bash
+pip install mailwise
+```
+
+Or from source:
+
+```bash
+git clone https://github.com/PetrGuan/MailWise.git
+cd MailWise
+pip install -e .
+```
+
+### Configure
+
+The easiest way to get started:
+
+```bash
+mailwise init
+```
+
+This will walk you through setting up your EML directory, adding expert engineers, and verifying the setup.
+
+Or configure manually:
+
+```bash
+cp config.example.yaml config.yaml
+```
+
+Edit `config.yaml` with your settings:
+
+```yaml
+eml_directory: /path/to/your/eml/files
+database: data/index.db
+markdown_directory: markdown
+embedding_model: all-MiniLM-L6-v2
+expert_boost: 1.5
+
+experts:
+  - email: senior.dev@company.com
+    name: Jane Doe
+```
+
+### Usage
+
+```bash
+# Index your emails (incremental — only processes new/changed files)
+mailwise index
+
+# Search for similar past issues
+mailwise search "sync failure after folder migration"
+
+# Search with previews
+mailwise search "calendar not updating" --show-body
+
+# Only show expert replies
+mailwise search "deleted emails reappear" --expert-only
+
+# Deep analysis — Claude reasons over similar expert threads
+mailwise analyze "User reports emails moved to local folder keep reappearing in Inbox"
+
+# View full markdown of a specific email thread
+mailwise show 42
+
+# Check index stats
+mailwise stats
+```
+
+### Managing experts
+
+```bash
+# Add an expert
+mailwise experts add engineer@company.com --name "Jane Doe"
+
+# List all experts
+mailwise experts list
+
+# Remove an expert
+mailwise experts remove engineer@company.com
+```
+
+## How it works
+
+```
+EML files → Parser → Markdown + Embeddings → SQLite index
+                                                    ↓
+                              Query → Semantic search → Top matches
+                                                            ↓
+                                          Claude (via RAG) → Expert-informed analysis
+```
+
+1. **Parse**: EML files are parsed in parallel and threads are split into individual replies using Outlook-style `From:/Sent:` delimiters
+2. **Clean**: Microsoft SafeLinks are unwrapped, mailto artifacts are removed
+3. **Markdown**: Each thread becomes a structured markdown file with `[Expert]` tags on replies from your designated engineers
+4. **Embed**: Each reply is embedded using `all-MiniLM-L6-v2` (runs locally, no API calls)
+5. **Index**: Embeddings and metadata are stored in SQLite for fast retrieval
+6. **Search**: Cosine similarity with expert score boosting finds relevant past issues
+7. **Analyze**: Top matches are fed to Claude (via Claude Code CLI) with a system prompt that focuses on expert reasoning patterns
+
+## Performance
+
+Designed for large mailboxes (25K+ emails, 16GB+):
+
+| Operation | Performance |
+|---|---|
+| Incremental check (no changes) | ~2-3s for 25K files (stat-based, no file reads) |
+| Full index | ~5-10 min (parallel parsing + batch embedding) |
+| Search query | <100ms (single matrix multiply over 100K+ vectors) |
+| RAG analysis | ~10-20s (retrieval + Claude response) |
+
+Key optimizations:
+- **Two-phase change detection**: mtime+size stat check before SHA256 hashing
+- **Parallel EML parsing**: multiprocessing with configurable workers
+- **Batch embedding**: pre-computed offset arrays, no O(n²) lookups
+- **Optimized search**: loads only embedding BLOBs into contiguous numpy array; fetches metadata only for top-k results
+- **SQLite tuning**: WAL journal, 64MB cache, 256MB mmap, batch inserts via `executemany`
+
+## Architecture
+
+```
+src/email_issue_indexer/
+├── cli.py          # Click-based CLI
+├── parser.py       # EML parsing + thread splitting (parallel-safe)
+├── markdown.py     # Markdown conversion with expert tags
+├── safelinks.py    # Microsoft SafeLinks URL cleaning
+├── embeddings.py   # sentence-transformers embeddings + vector search
+├── store.py        # SQLite storage layer (performance-tuned)
+├── indexer.py      # Parallel batch orchestrator with progress tracking
+├── search.py       # Optimized similarity search with expert boosting
+└── rag.py          # RAG layer using Claude Code CLI
+```
+
+## Privacy
+
+All processing is local:
+- Embeddings run on your machine (no data sent to any API for indexing)
+- Email content stays in your local SQLite database and markdown files
+- The `analyze` command sends relevant thread excerpts to Claude — same as chatting in Claude Code
+
+Your `config.yaml`, `emails/`, `data/`, and `markdown/` directories are gitignored by default. Only `config.example.yaml` (with no real data) is committed. A pre-commit hook (`scripts/install-hooks.sh`) scans for accidental PII leaks.
+
+## License
+
+MIT

mailwise-0.1.0/pyproject.toml

@@ -0,0 +1,50 @@
+[build-system]
+requires = ["setuptools>=61.0"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "mailwise"
+version = "0.1.0"
+description = "Turn email threads into a searchable knowledge base. Parse EML files, index with embeddings, and use RAG to learn how your best engineers analyze issues."
+readme = "README.md"
+license = "MIT"
+requires-python = ">=3.10"
+authors = [
+    { name = "Petr" },
+]
+keywords = ["email", "eml", "rag", "embeddings", "issue-tracking", "knowledge-base"]
+classifiers = [
+    "Development Status :: 3 - Alpha",
+    "Intended Audience :: Developers",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Topic :: Communications :: Email",
+    "Topic :: Text Processing :: Indexing",
+]
+dependencies = [
+    "sentence-transformers>=2.2.0",
+    "numpy>=1.24.0",
+    "click>=8.0",
+    "pyyaml>=6.0",
+]
+
+[project.urls]
+Homepage = "https://github.com/PetrGuan/MailWise"
+Issues = "https://github.com/PetrGuan/MailWise/issues"
+
+[project.scripts]
+mailwise = "email_issue_indexer.cli:cli"
+
+[tool.setuptools.packages.find]
+where = ["src"]
+
+[project.optional-dependencies]
+dev = [
+    "pytest>=7.0",
+    "pytest-cov>=4.0",
+]
+
+[tool.pytest.ini_options]
+testpaths = ["tests"]

mailwise-0.1.0/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

mailwise-0.1.0/setup.py

@@ -0,0 +1,2 @@
+from setuptools import setup
+setup()

mailwise-0.1.0/src/email_issue_indexer/__init__.py

@@ -0,0 +1,2 @@
+"""Email Issue Indexer - Parse, index, and search email issue threads."""
+__version__ = "0.1.0"