agentleFS
QuixiAIQuixiAI/Hexis/skills/installed/inbox-triage/SKILL.md

inbox-triage

Skill623 starsChanged 10 months agoMITView on GitHub

Review an inbox, propose a safe triage plan, and apply only the email actions the user approves

52 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use aeon-update, auto-workflow and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • aaronjmarsSkill

    aeon-update

    Pull framework updates from the upstream Aeon repo into this instance - 3-way merges canon's new commits into a PR, never clobbering operator config.

    Also in 4 projects that use this

  • aaronjmarsSkill

    auto-workflow

    Two-mode aeon.yml workflow builder - analyze inspects URLs and emits a tiered, signal-verified skill-enablement plan plus an aeon.yml diff; enable flips slugs to enabled:true and opens a PR.

    Also in 4 projects that use this

  • +10 more from aaronjmars

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 5 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 44 edited versions we read keep these.

  • 5

    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:

    List before reading: Use message metadata to identify likely high-value threads.

    line 19
  2. Distinctive:

    Apply the chosen batch: Only after the user chooses, apply labels, archive, or mark spam to the exact messages approved.

    line 22
  3. Distinctive:

    Never permanently delete during ordinary triage.

    line 28
  4. Distinctive:

    Never infer that a newsletter is spam or that silence means approval.

    line 29
  5. Distinctive:

    If the account is disconnected or a provider action is unavailable, preserve the proposed plan and give the exact connection or manual next …

    line 32

“Common” is measured: kept by at least 6 in 10 of the 44 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

Inbox Triage

Reduce inbox load while keeping classification, replies, and destructive actions under the user's control. Use email-digest instead when the user wants only a summary or memory ingestion.

Workflow

  1. Set scope. Confirm the account when more than one is connected, then honor the requested folder, query, date range, or message limit. Default to a small recent inbox sample rather than scanning all mail.
  2. List before reading. Use message metadata to identify likely high-value threads. Read only enough candidate messages to classify them accurately; do not bulk-read the whole mailbox.
  3. Classify with reasons. Group messages into: reply needed, task or deadline, waiting/reference, safe archive, likely spam, and uncertain. Include message IDs or unambiguous sender/subject labels so the proposed actions are inspectable.
  4. Show the plan first. Present counts and the proposed action for each thread. Keep uncertain messages out of action batches and ask about them.
  5. Apply the chosen batch. Only after the user chooses, apply labels, archive, or mark spam to the exact messages approved. Report successes and failures per message.
  6. Handle replies separately. Draft a reply, show recipients, subject, and body, then send only after explicit approval. A request to triage is not permission to send.
  7. Capture tasks deliberately. Suggest tasks for real commitments. Create them in the local backlog, Todoist, or Asana only after the user chooses the destination and content.

Control and Failure Boundaries

  • Never permanently delete during ordinary triage. Use trash only after an explicit request; permanent deletion requires unmistakable, message-specific authorization.
  • Never infer that a newsletter is spam or that silence means approval. Archive, spam, labels, replies, and tasks are separate choices.
  • Respect the configured Gmail memory policy. Do not store email-derived content when it says forget; ask before storage when it says ask; summarize rather than storing raw bodies.
  • Gmail connection is not autonomous permission. This skill is chat-only and runs because the user asked now.
  • If the account is disconnected or a provider action is unavailable, preserve the proposed plan and give the exact connection or manual next step. Do not discard the useful read-only work.

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.