Claude Code Chief of Staff plugin

A plugin that turns Claude Code into a personal chief of staff with persistent memory, a morning brief and a weekly review

  • 92 ★ on the repository
  • Last commit
  • MIT
Plugin

Extended access

Why this level

  • The plugin creates personal memory files and the brief reads calendar, email and messages through connected sources.
All reasons and checks

loganhc-09/claude-chief-of-staff

Install

Works in

In your terminal, with SkillFoxx CLI

npx skillfoxx add plugins/claude-code-chief-of-staff

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • The plugin creates personal memory files and the brief reads calendar, email and messages through connected sources.

This is third-party code. Review the repository files before installing.

What it does

The chief-of-staff plugin builds a persistent personal organization layer around Claude Code. `/chief-of-staff:setup` briefly interviews you (what you do, what you work on, how the assistant should behave, your writing voice, who matters) and creates `CHIEF_OF_STAFF.md`, `MEMORY.md`, a task list and a brief template in `~/chief-of-staff/`. All of it is plain Markdown on your machine. `/chief-of-staff:brief` builds a morning or evening brief: what is on the calendar, what is overdue, what to prepare for, one thing worth attention up top. Before calling a task overdue the agent checks it against sent mail, notes and the calendar. `/chief-of-staff:loops` runs a weekly review: for every open loop the agent recommends ship, block or kill. The repository also describes the author's wider setup: launchd background jobs pull email, calendar and meeting transcripts, results arrive by email and urgent items by iMessage. Sending messages and publishing content stay with the human.

Who it is for. Managers and solo professionals with many meetings, emails and commitments who want an assistant that remembers context between Claude Code sessions.

Good fit when

  • Every morning you need a brief: the day's events, overdue promises and a couple of things worth preparing for
  • On Friday you want to go through everything in flight and decide on each item instead of adding new ones
  • The assistant should remember your writing style, the people around you and rules like never sending without showing you
  • You need a `CLAUDE.md` template and a memory scheme to start from when building your own system
  • Meeting transcripts and incoming email should become facts and a task list through background scripts

Not a fit when

  • You want one ready service with an interface: this is Markdown files and scripts you set up yourself
  • You do not want an agent to reach email, calendar and transcripts: the brief relies on connected sources
  • You are not on a Mac and want to copy the background part: the author's examples assume launchd jobs

Example request

/chief-of-staff:brief — build my morning brief: today's events, overdue promises and what needs preparing.

Limitations

The plugin installs only the core: memory, the brief and the review. The background part (the `briefing.py` and `meeting_processor.py` scripts and launchd jobs) sits in the repository as an example and is assembled by hand on a Mac. Email, calendar and messages need connected sources, and without them the brief relies on what the agent can see itself. The author runs on Claude Max at 200 dollars a month and notes that most of the setup also works on the 100 dollar plan. Keep personal context outside git repositories: the setup command warns if the chosen folder sits inside one. The system sprawls: by September the author had 199 scripts and 52 scheduled jobs.

How to disable. Remove the plugin through `/plugin`. The `~/chief-of-staff/` folder with your memory and task list stays on disk and is deleted by hand, and launchd jobs are unloaded separately.

FAQ

How do I install the plugin?

Run `/plugin marketplace add loganhc-09/claude-chief-of-staff` and `/plugin install chief-of-staff@chief-of-staff`, restart Claude Code and run `/chief-of-staff:setup`.

Which commands does the plugin add?

`/chief-of-staff:setup` creates the system, `/chief-of-staff:brief` builds a morning or evening brief and `/chief-of-staff:loops` runs a weekly review of open loops.

Where is my data stored?

In plain Markdown files in `~/chief-of-staff/` unless you pick another folder. The author advises keeping it outside any repository.

What does the human decide and what does the agent do?

The agent processes transcripts, syncs data and prepares briefs. It only drafts replies to email. Sending messages, publishing and strategic decisions stay with you.

Do I need a Claude Max subscription?

The author uses Max at 200 dollars a month and says most of the setup works on Max 5x at 100 dollars. Plan and cost advice is in `getting-started.md`.

Discussion

No questions or reviews yet.

Plugins

A plugin for Claude Code and Codex: department folders with rules and 35 skills to run a one-person company through AI agents

PluginExtended access244Repository stars

A Claude Desktop template where nine scheduled tasks build a morning brief from email, WhatsApp and calendar.

WorkflowBroad access122Repository stars

An Obsidian vault template for Claude Code: /daily, /weekly, /monthly and /project commands link three-year goals to today's tasks

SkillBroad access1.9KRepository stars
Needs a VPN

A Claude Code plugin that catches your corrections in chat and, after review, writes them into CLAUDE.md and AGENTS.md

PluginExtended access1.9KRepository stars

Claudia turns Claude Code or Codex into an assistant that keeps people and commitments in SQLite and an Obsidian wiki.

PluginExtended access301Repository stars

Eight Claude Code skills for Obsidian: Granola notes, NotebookLM imports, TaskNotes tasks, recall of past sessions, chat export

PluginExtended access552Repository stars

README in short

The README covers plugin installation, the three layers of the system (knowledge, live work, autonomous jobs), a handoff model with three modes and the author's mistakes over a year, such as silent job failures and script sprawl.

README badge

Report a problem