agentleFS
mediajunkiemediajunkie/piper-morgan-product/.claude/skills/deliver-mail/SKILL.md

deliver-mail

Skill8 starsChanged 11 days agoApache-2.0View on GitHub

RETIRED 2026-06-19 (post-migration, #1259). The web↔code PM mail-shuttle this skill described no longer exists. SEND mail via push-to-ref (scripts/mail-send.sh — see CLAUDE.md "mailbox workflow"); RECEIVE mail via the check-mailbox skill. Kept only as a redirect — do not use it for mail operations.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    After the June 2026 migration wave the whole cohort runs on Claude Code.

    line 25
  2. Very specific:

    Check mailboxes/{your-slug}/inbox/, read each memo, move it to read/, respond where Response-Requested: yes, and note action items in your …

    line 36
  3. Very specific:

    Write the memo + any CC copies + your sent/ mirror, and do any inbox/ → read/ moves — just write/mv the files at their mailboxes/… paths.

    line 45

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “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

deliver-mail — RETIRED

This skill is retired. It does not describe how mail works anymore. It is kept as a redirect so old invocations of /deliver-mail land somewhere useful instead of a dead end.

Why it was retired

deliver-mail automated the chat-era web↔code PM-shuttle: a human PM downloaded memos that "web agents" (roles running on claude.ai) had written, dropped them in mailboxes/incoming/, and this skill routed them into code-side inboxes (Phase 1 Ingest); then it walked the PM through hand-delivering code-side memos back to web agents (Phase 2 Outbound Audit).

After the June 2026 migration wave the whole cohort runs on Claude Code. There are no web agents and no PM-download shuttle, so Phases 1 and 2 have no referent. The old skill also claimed "Mailboxes are gitignored: …Nothing is committed to git" — that is now flatly wrong: mailboxes are git-tracked and every mail op lands on origin/main (see below).

Where its purpose lives now

The mail lifecycle is fully covered by two current surfaces. Use these directly:

To RECEIVE / triage your inbox → the check-mailbox skill

Check mailboxes/{your-slug}/inbox/, read each memo, move it to read/, respond where Response-Requested: yes, and note action items in your session log. (The recipient also owns regenerating their own inbox MANIFEST.md.)

To SEND / route a memo → push-to-ref via scripts/mail-send.sh

Canonical procedure: CLAUDE.md → "The mailbox workflow (most-frequent case) — push-to-ref via mail-send.sh" (updated 2026-06-19, #1259). In short, from your own worktree:

  1. Write the memo + any CC copies + your sent/ mirror, and do any inbox/ → read/ moves — just write/mv the files at their mailboxes/… paths. Do not git add/commit them by hand.
  2. Send, passing every changed path explicitly (new files and the inbox side of a move):

    scripts/mail-send.sh "mail({role}): {subject}" \
        mailboxes/{recipient}/inbox/{memo}.md \
        mailboxes/{you}/sent/{memo}.md

mail-send.sh (v3) builds the commit as a git object on top of origin/main (commit-tree via a throwaway index) and pushes it straight to main. It never touches a shared working tree or the local main ref, so concurrent agents can't sweep or strand each other. On a non-fast-forward it rebuilds on the new tip and retries automatically. After sending, the files sit uncommitted on your worktree branch and reconcile on your next sync — that's expected.

The old bridge dance (stash → checkout main → git add mailboxes/ → push → switch back) is also retired — do not do it. The check-branch.sh PreToolUse hook remains as the backstop for any interactive mail commit on a non-main branch (commit-tree isn't git commit, so mail-send.sh doesn't trip it).

Reference

  • Send: CLAUDE.md → "The mailbox workflow (most-frequent case) — push-to-ref via mail-send.sh"
  • Receive: the check-mailbox skill
  • Full rule set: docs/internal/operations/branch-worktree-mailbox-discipline.md
  • Why push-to-ref: docs/internal/operations/mailbox-bridge-transparency-design-2026-06-16.md (#1259)
  • Routing reference: mailboxes/DIRECTORY.md (canonical slug → role map)

(This skill may be hard-deleted once the cohort is confident no workflow still invokes /deliver-mail.)

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.