← 返回 Skills 市场
petrguan

MailWise

作者 PetrGuan1995 · GitHub ↗ · v1.0.2 · MIT-0
cross-platform ⚠ suspicious
299
总下载
1
收藏
0
当前安装
3
版本数
在 OpenClaw 中安装
/install mailwise
功能描述
Search and analyze email issue threads from a local knowledge base. Use when the user asks about past bugs, incidents, or wants to find how experienced engin...
使用说明 (SKILL.md)

MailWise — Your Team's Email Memory

Stop re-investigating issues your team already solved. MailWise turns your email threads into a searchable knowledge base — index thousands of EML files locally, tag expert engineers, and use RAG to instantly surface how similar issues were debugged and resolved in the past.

"Have we seen this before?" → MailWise searches 25,000+ emails in under a second and shows you exactly what your best engineers said about it.

What you get

Semantic search — not keyword matching, but meaning-based retrieval across your entire email archive:

$ mailwise search "calendar sync fails after folder migration" --show-body

  #1 [0.89] ★Expert  RE: Outlook calendar not syncing after mailbox move
     From: [email protected] (2024-11-15)
     "The root cause is a stale folder-id cache. Clear HxCalendarSync
      and restart the sync service..."

  #2 [0.84]          RE: Sync failure post-migration — Mac clients
     From: [email protected] (2024-10-22)
     "We've seen this pattern before. Check if the migration tool
      preserved the folder GUIDs..."

  #3 [0.81] ★Expert  RE: Calendar items disappearing after move
     From: [email protected] (2024-09-08)
     "This is the same class of bug as CASE-4521. The fix is..."

RAG-powered analysis — Claude reads the most relevant past threads and synthesizes debugging guidance:

$ mailwise analyze "User reports calendar events vanish after moving to a new folder"

  Based on 5 similar past issues, here's what your team has found:
  ● Root cause pattern: stale folder-id cache after migration (3 of 5 cases)
  ● Recommended first step: check HxCalendarSync logs for GUID mismatch
  ● Expert consensus: clearing the sync cache resolves ~80% of these cases
  ● Edge case: if GUIDs were not preserved during migration, a full re-sync is needed

Get started in 3 steps

pip install mailwise        # 1. Install from PyPI
mailwise init               # 2. Interactive setup wizard — creates config, sets directories
mailwise index              # 3. Index your EML files (incremental, safe to re-run)

That's it. Now search with mailwise search "..." or get AI analysis with mailwise analyze "...".

Commands

Search for similar issues

mailwise search "describe the issue here" --show-body
  • --show-body — show a preview of each matching message
  • --expert-only — only show replies from expert engineers
  • -k N — number of results (default: 10)

Deep analysis with RAG

mailwise analyze "paste full bug report or issue description here"
  • -k N — number of similar issues to feed to Claude (default: 5)

Tip: Paste the FULL bug report — error codes, logs, environment details produce much better matches.

mailwise analyze "$(cat bug_report.txt)" -k 10

View a full email thread

mailwise show \x3CEMAIL_ID>

Check index status

mailwise stats

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 \x3CID> to read full threads of interest
  5. Summarize findings: root cause patterns, debugging steps, and next actions

Privacy & data handling

  • Local-only commands (index, search, show, stats, experts): Everything runs on your machine. Embeddings generated locally via sentence-transformers. No data sent anywhere.
  • External LLM command (analyze): Sends email excerpts to the Anthropic API via Claude Code CLI. Do not use on sensitive emails unless your org's data policy permits it.

Authentication

  • The analyze command requires Claude Code CLI — run claude once to authenticate, or set ANTHROPIC_API_KEY
  • No auth needed for local commands (index, search, show, stats, experts)
安全使用建议
This skill appears coherent, but take these precautions before installing: 1) Inspect and verify the PyPI package owner/version (pip installs run third-party code). Install in an isolated virtualenv/container and check the package source (or GitHub repo) if possible. 2) Do not use mailwise analyze on sensitive or regulated emails unless your org permits sending excerpts to Anthropic; prefer authenticating the Claude CLI interactively (avoids storing an API key in env). 3) Review the config.yaml and indexed directory choices created by mailwise init to ensure only intended EML files are indexed. 4) Monitor network activity when running 'analyze' to confirm only calls to Anthropic occur. 5) If you need higher assurance, request the package source or a checksum from the publisher and audit it before deployment.
功能分析
Type: OpenClaw Skill Name: mailwise Version: 1.0.2 The MailWise skill bundle (SKILL.md, reference.md) provides semantic search and RAG-based analysis for local email archives. It requires risky capabilities such as local file system access to index EML files and network access to transmit data to the Anthropic API for analysis. While these actions are clearly aligned with the tool's stated purpose and are transparently documented, the inherent risks of processing sensitive email data and the reliance on an external PyPI package (mailwise) meet the criteria for a suspicious classification under the provided guidelines.
能力评估
Purpose & Capability
Name/description (search & analyze email threads) align with declared requirements: a mailwise CLI for local indexing/search and an optional 'claude' CLI (or ANTHROPIC_API_KEY) for the analyze command. Requiring 'claude' for external LLM analysis is coherent with the documented feature.
Instruction Scope
SKILL.md confines local commands (index, search, show, stats, experts) to on-disk EML files and local embeddings. The only external network activity is the optional 'analyze' command, which sends selected excerpts to Anthropic via the Claude CLI. The docs explicitly warn about not using analyze on sensitive emails — this is within scope but is an important privacy caveat because users are encouraged to paste full bug reports/logs which may contain sensitive data.
Install Mechanism
Install spec is a PyPI package (pip/uv -> mailwise) which is proportionate for a Python CLI tool. Pip installs run third-party code (package/setup scripts and dependencies), so installing from PyPI carries the usual moderate risk; the skill itself includes only docs (no code files) so the scanner had nothing to audit locally.
Credentials
No required env vars. ANTHROPIC_API_KEY is optional and only needed for the external 'analyze' command; that is proportional and documented. There are no unrelated credentials/config paths requested.
Persistence & Privilege
The skill does not request 'always' or elevated privileges. The interactive setup writes a local config.yaml and directories (expected behavior for indexing). No indication it modifies other skills or system-wide agent settings.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install mailwise
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /mailwise 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.2
**MailWise 1.0.2 — Now installable from PyPI, with streamlined setup and updated docs.** - Switched default installation to PyPI (`pip install mailwise`) for easier setup. - Added `mailwise init` interactive setup wizard for faster first-time configuration. - Updated instructions and description to reflect simpler workflow (no git clone required). - Clarified privacy and authentication details. - Improved documentation with clearer examples and usage tips.
v1.0.1
- Documented privacy and data handling for all commands, highlighting that only the analyze command sends data externally (to Anthropic via Claude). - Clarified authentication requirements, including use of the `ANTHROPIC_API_KEY` environment variable or Claude CLI login for analyze. - Explicitly noted that all local commands (search, index, show, stats, experts) process data only on your machine. - No code changes; SKILL.md documentation improvements only.
v1.0.0
- Initial release of MailWise skill for searching and analyzing email issue threads from a local knowledge base. - Supports searching for similar past incidents, bug resolutions, and expert replies using command-line tools. - Integrates Retrieval-Augmented Generation (RAG) analysis via Claude for deep synthesis of root cause patterns and debugging tactics. - Allows indexing new .eml emails, managing expert engineer lists, and tracking coverage stats. - All operations are performed locally; no email data is sent to external services.
元数据
Slug mailwise
版本 1.0.2
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 3
常见问题

MailWise 是什么?

Search and analyze email issue threads from a local knowledge base. Use when the user asks about past bugs, incidents, or wants to find how experienced engin... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 299 次。

如何安装 MailWise?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install mailwise」即可一键安装,无需额外配置。

MailWise 是免费的吗?

是的,MailWise 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

MailWise 支持哪些平台?

MailWise 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 MailWise?

由 PetrGuan1995(@petrguan)开发并维护,当前版本 v1.0.2。

💬 留言讨论