MailWise — Email Issue Knowledge Base

Search and analyze your team's email issue threads. MailWise indexes EML files, tags expert engineers' replies, and uses RAG to help you learn from how experienced engineers investigated and resolved past issues.

When to use this skill

• User asks about similar past issues or bugs
• User wants to know how an issue was resolved before
• User pastes a bug report and wants analysis based on team history
• User asks about root cause patterns or debugging approaches
• User wants to find what expert engineers said about a topic

Commands

Search for similar issues

Find past issues matching a description. Returns ranked results with expert tags.

mailwise search "describe the issue here" --show-body

Options: - --show-body — show a preview of each matching message - --expert-only — only show replies from expert engineers - -k N — number of results (default: 10)

Example:

mailwise search "email sync failure after folder migration" --show-body
mailwise search "calendar not updating on Mac" --expert-only -k 5

Deep analysis with RAG

Retrieve similar past issues and ask Claude to synthesize expert insights, root cause patterns, and debugging approaches.

mailwise analyze "paste full bug report or issue description here"

Options: - -k N — number of similar issues to feed to Claude (default: 5, increase for broader analysis)

Tip: For best results, paste the FULL bug report content, not just a title. More context (error codes, logs, environment details) produces better matches.

Example:

mailwise analyze "User on Mac Outlook moves emails from Inbox to local folder. Emails disappear but reappear after 30 minutes. No DeleteMessage in HxLogs."
mailwise analyze "$(cat bug_report.txt)" -k 10

View a full email thread

After finding an issue via search, view the complete parsed thread with all replies and expert tags.

mailwise show <EMAIL_ID>

Check index status

mailwise stats

Shows: total indexed emails, thread messages, expert messages, expert coverage percentage.

Index new emails

When the user adds new EML files or wants to refresh the index:

mailwise index

This is incremental — only processes new or changed files. Safe to run repeatedly.

Manage expert engineers

mailwise experts list
mailwise experts add [email protected] --name "Jane Doe"
mailwise experts remove [email protected]

Expert engineers' replies get [Expert] tags and boosted scores in search results.

Typical workflow

1. User describes an issue or pastes a bug report
2. Run mailwise search "..." --show-body to find similar past issues
3. If promising results found, run mailwise analyze "..." for deep RAG analysis
4. Use mailwise show <ID> to read full threads of interest
5. Summarize findings: root cause patterns, debugging steps, and next actions

Setup requirements

Before first use, the user needs to:

1. Install MailWise: bash pip install git+https://github.com/PetrGuan/MailWise.git
2. Clone the repo for config files: bash git clone https://github.com/PetrGuan/MailWise.git && cd MailWise cp config.example.yaml config.yaml
3. Edit config.yaml: set eml_directory and add expert engineers
4. Put .eml files in the configured directory
5. Run mailwise index to build the initial index

Privacy & data handling

• Local-only commands (index, search, show, stats, experts): All processing runs entirely on your machine. Embeddings are generated locally via sentence-transformers. No data is sent to external services.
• External LLM command (analyze): This command sends email excerpts (search results matching your query) to the Anthropic API via the Claude Code CLI (claude --print). Treat any content passed to analyze as potentially transmitted to Anthropic's servers. Do not use analyze on sensitive emails unless your organization's data policy permits it. You can use all other commands without any external API calls.

Authentication requirements

The analyze command requires Claude Code CLI to be installed and authenticated:

• Binary: claude (Claude Code CLI)
• Auth: Run claude once to authenticate via Anthropic Console, or set the ANTHROPIC_API_KEY environment variable
• No auth needed for local-only commands (index, search, show, stats, experts)

Other notes

• The index is stored in a local SQLite database
• Markdown versions of all parsed threads are written to the markdown/ directory