The latest articles on DEV Community by Don Yin (@donyin). https://dev.to/donyin https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3831563%2F12197376-1954-4ac7-86bc-3f614e987bcb.png https://dev.to/donyin en Don Yin Wed, 18 Mar 2026 14:11:05 +0000 https://dev.to/donyin/apple-mail-an-agent-skill-that-gives-ai-assistants-full-control-over-macos-mailapp-188n https://dev.to/donyin/apple-mail-an-agent-skill-that-gives-ai-assistants-full-control-over-macos-mailapp-188n <p>I open-sourced an agent skill that gives AI coding assistants full control over macOS Mail.app: read, search, compose, reply, forward, move, delete, and label emails without leaving the editor.</p> <p>It ships as a self-contained skill directory. Drop it into your agent's skill folder (e.g. <code>.cursor/skills/</code>) and the agent can triage your inbox, draft replies, and organise mail on your behalf. It works with Cursor, Claude Code, or any agent that can invoke shell commands.</p> <h2> Why? </h2> <p>Scripting Bridge calls to Mail.app are slow (seconds per message). This skill reads <code>.emlx</code> files directly from disk (~5 ms per message) and maintains a SQLite FTS5 search index for fast full-text search across body, subject, and sender.</p> <h2> Highlights </h2> <ul> <li> <strong>Disk-first reads</strong>: parses <code>.emlx</code> files directly (~5 ms per message vs seconds via the scripting bridge)</li> <li> <strong>FTS5 search</strong>: SQLite full-text search index across body, subject, and sender</li> <li> <strong>Exchange-safe deletes</strong>: uses RFC Message-ID so it works reliably with Exchange accounts</li> <li> <strong>Draft-first sending</strong>: compose and reply always create drafts first; nothing sends without explicit confirmation</li> <li> <strong>Dry-run support</strong>: destructive operations (delete, amend-subject) support <code>--dry-run</code> to preview changes</li> </ul> <h2> Commands </h2> <p>All commands go through <code>scripts/mail.sh</code> and return structured JSON:</p> <p><code>list-accounts</code>, <code>list-folders</code>, <code>list-recent</code>, <code>list-emails</code>, <code>list-drafts</code>, <code>read-email</code>, <code>search</code>, <code>compose-draft</code>, <code>amend-draft</code>, <code>send-draft</code>, <code>reply-draft</code>, <code>forward-draft</code>, <code>delete-email</code>, <code>delete-draft</code>, <code>move-email</code>, <code>amend-subject</code>, <code>add-label</code>, <code>build-index</code></p> <h2> Setup </h2> <div class="highlight js-code-highlight"> <pre class="highlight shell"><code><span class="nb">cp</span> <span class="nt">-r</span> apple-mail /path/to/project/.cursor/skills/ </code></pre> </div> <p>On first use, <code>mail.sh</code> auto-creates a micromamba environment with Python 3.11 and installs dependencies.</p> <p><strong>Requirements:</strong> macOS with Mail.app configured, micromamba, Full Disk Access for Terminal.</p> <h2> Links </h2> <ul> <li> <strong>GitHub:</strong> <a href="https://github.com/Don-Yin/apple-mail" rel="noopener noreferrer">Don-Yin/apple-mail</a> </li> <li> <strong>License:</strong> MIT</li> </ul> <p>Feedback and PRs welcome!</p> opensource ai productivity python