agentleFS
fabiohuebelfabiohuebel/claude-mac-bridge/skills/mail/SKILL.md

mail

Skill0 starsChanged 5 months agoMITView on GitHub

Triage your Apple Mail from a Mac. Pull recent messages, classify them, and flag the ones that need action. Use to check email, see what needs a reply, or run a recurring inbox triage (wrap in /loop).

8 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 5 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do not flag mailing-list/newsletter noise as Action.

    line 30

“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Mail (Apple Mail triage)

Reads and flags Apple Mail through AppleScript helper scripts in scripts/. Mail.app aggregates all your accounts in one place, so this works across every account you have configured there. Mac-only. Read + flag only: it never sends, deletes, archives, or marks read.

Personalize first: set DEFAULT_ACCOUNTS at the top of scripts/list.sh to your Mail.app account name(s). Use the exact "Description" shown under Mail > Settings > Accounts (e.g. iCloud, Gmail, Work). Account names like Work below are placeholders.

On invocation (triage routine)

  1. Pull recent mail. bash ~/.claude/skills/mail/scripts/list.sh (default account, 10 most recent). For a wider sweep pass an account name: list.sh Work, list.sh iCloud. Output marks [UNREAD] vs [read].
  2. Classify each unread/recent message into:
    • 🔴 Action: needs a reply or a task (a real person asking something, admin with a deadline, bills, anything time-sensitive).
    • 🟠 FYI: informational, worth knowing, no action (announcements, receipts).
    • ⚪️ Noise: mailing lists and newsletters. Summarize as a count, do not itemize.
  3. Flag the Action items (reversible): bash ~/.claude/skills/mail/scripts/flag.sh "<account>" "<subject substring>" on red. Use red for must-do, orange for a softer "should reply". Unflag with ... off.
  4. Present a tight triage, not a dump: a 🔴 Action list (sender + what they want + suggested next step), a short 🟠 FYI list, and "+N list-serv/newsletter items skipped". End with the single most important thing to handle, and offer to draft a reply (draft only, see guardrails).

Recurring use

Wrap the routine in a loop: /loop 45m /mail (runs while the Mac and a session are on). Each pass re-triages and flags new Action mail. Flags show up in Mail.app's Flagged smart mailbox and sync to iPhone.

Scripts

  • list.sh [account] [count]: fast bulk read (avoids the slow unified-inbox count); resilient inbox resolver. Default account set is editable at the top of the script (DEFAULT_ACCOUNTS).
  • flag.sh "<account>" "<subject substring>" [on|off] [color]: flag/unflag the first matching message (within the 40 most recent). Colors: red orange yellow green blue purple grey.

Guardrails

  • Read + flag only. Drafting a reply is fine (tell the user, or use a Gmail connector's create_draft if available), but never auto-send, delete, archive, or mark-read without explicit say-so.
  • Do not flag mailing-list/newsletter noise as Action.
  • Account names must match Mail.app exactly, including any punctuation or spaces. Copy the name as shown under Mail > Settings > Accounts.
  • Mail's unified-inbox count is very slow: always read per-account with the bounded slice (the scripts already do this). Do not iterate the unified inbox.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.