Skip to content

README.en.md

Latest commit

 

History

History
144 lines (102 loc) · 6.05 KB

README.en.md

File metadata and controls

144 lines (102 loc) · 6.05 KB
Raw

execution-os

Version License Claude

🌍 Français · English

A personal execution system for Claude: turns intent into results, with 4 modular modules (sprint planning, deep work, weekly ritual, accountability).

What it does

Most people confuse planning with execution. They build to-do lists that never get finished. execution-os applies 4 hard principles:

  1. A sprint has a single promise — not 10 goals, ONE central deliverable
  2. Time is the constraint, not the task — you allocate blocks, not intentions
  3. The retro beats the plan — the weekly ritual matters more than the beauty of the initial planning
  4. Accountability is brutal but fair — no sugar-coating when you drift

The 4 modules

Module When to use it
Sprint Planner Start a 1-2 week sprint, new project, new quarter
Deep Work Designer When you have 2-4h ahead and a big topic to move forward
Weekly Ritual Sunday evening / Monday morning, retro + weekly plan
Accountability Check When you drift, procrastinate, or question your commitments

Triggers

The skill triggers automatically when you say:

  • "Plan my week / my sprint" → Sprint Planner
  • "I have 3h ahead, what should I do" → Deep Work Designer
  • "I'm doing my retro", "Sunday evening" → Weekly Ritual
  • "I'm drifting", "I'm procrastinating on X", "I can't stick to my plan" → Accountability Check

Complementary skills (full ecosystem)

The 5 answer different questions:

  • thinking-os — WHAT to do (thinking frame)
  • execution-os — HOW to do it (cadence)
  • learning-os — HOW TO LEARN it (skill building)
  • strategy-intel — WHERE & AGAINST WHOM to do it (positioning)
  • alignment-os — WHY do it (life/values coherence)

Installation

Claude.ai

  1. Download or clone this folder
  2. Upload to /mnt/skills/user/execution-os/
  3. The skill triggers automatically on matching prompts

Claude Code

cp -r execution-os ~/.claude/skills/

Claude API

Package as .skill and upload via the /skills endpoint.

Examples

Sprint planning

"I'm launching cohort 2 of my program in 3 weeks — I need to finish the curriculum, recruit 10 mentors, validate 15 startups."

→ The skill refuses the 4 deliverables, forces picking ONE (the 15 startups), produces a calendarized plan with a protected Thursday buffer.

Deep work

"I have 3h ahead of me. I want to move forward on my commercial strategy."

→ The skill forces an observable deliverable ("2-page doc, 3 options compared"), breaks it into 45-50 min blocks with breaks, enforces the wrap-up.

Weekly ritual

"This week's review: I didn't finish my main deliverable but I did a lot of other stuff."

→ The skill refuses the "but", quantifies the delivery rate against the central promise (0%), identifies the avoidance pattern, produces a plan attacking that pattern.

Accountability

"It's been 3 weeks that I've been saying I'll write a book. I haven't written anything. Maybe it's not the right moment?"

→ The skill refuses the escape, asks the 4 hard questions, distinguishes obstacle from internal resistance, produces a clear decision (either re-engage at 30% scope, or explicit clean abandon).

Structure

execution-os/
├── SKILL.md                            # Metadata + router
├── README.md                           # French version
├── README.en.md                        # This file
├── modules/
│   ├── sprint-planner.md
│   ├── deep-work-designer.md
│   ├── weekly-ritual.md
│   └── accountability-check.md
├── templates/                          # Reusable markdown templates
│   ├── sprint-plan-template.md
│   ├── weekly-ritual-template.md
│   └── deep-work-template.md
├── examples/                           # End-to-end examples
│   ├── sprint-example.md
│   ├── weekly-example.md
│   └── accountability-example.md
└── tests/
    └── test-cases.md                   # 11 validation cases

Philosophy

  • Progressive disclosure: the SKILL.md only contains the router. Each module is loaded on demand.
  • No free motivation: no "you can do it!", no coach quotes.
  • Reality priority: no over-optimization — a livable plan beats a perfect one.
  • Safety rule: if the user shows signs of burnout / exhaustion / distress, the skill stops and redirects to a real human support.

Limitations

  • Does not replace a therapist, a human coach, or a doctor
  • No gamification (no points, no streaks)
  • Time estimates remain estimates — the skill isn't magic
  • Not a good fit for tasks under 15 min (overhead > value)

Note: Content in French (modules, templates, examples) is currently the reference version. An English translation of module bodies is on the roadmap.

Roadmap

  • v1.0 (current): 4 modules + 3 templates + routing
  • v1.1: Monthly ritual (retro + plan at the monthly scale)
  • v1.2: Energy tracking (correlation energy / task types)
  • v1.3: Explicit composition with thinking-os (automatic redirection)
  • v2.0: Memory of past sprints to improve future estimates

License

Apache 2.0 — fork, adapt, republish.

Credits

Built with Claude as Skill Architect. Inspired by the Cal Newport tradition (deep work), Eisenhower (prioritization), Ray Dalio (principles), and Agile sprint practice.