rebelytics/one-skill-to-rule-them-all
Some data may be outdated — see provenance labels.
The meta-skill that builds and improves all your skills, including itself. Watches your work sessions (autonomous or human-led), captures patterns, corrections and judgement calls, and turns them into skill improvements and new skill candidates for your review. Practical application of the Augmented Expertise methodology. Open source: CC BY 4.0. observed · 2026-09-26
Health v2 · maintenance only
80/100
- Activity 96
- Release rhythm 96
- Longevity 17
How is this computed?
round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10) — computed 2026-10-21. Adoption (stars, forks) is never an input.
- gap_med: 6.5
- age_days: 247
- days_rel: 25
- days_push: 25
- n_releases_24m: 7
Adoption not part of the score
2869 stars · 286 forks observed · 2026-09-26
What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded
Task Observer is an open-source meta-skill (CC BY 4.0) that watches AI-assisted work sessions, captures corrections, preferences, and gaps, and turns them into improvements for your skill library — including improvements to itself. It also flags repeating patterns as candidates for new skills, implementing the Augmented Expertise methodology.
Use cases
- automatically improve my AI skills from corrections I make during sessions
- discover new skill candidates from repeating patterns in my work
- make my skill library self-improving
- capture cross-cutting principles that apply across all my skills
- set up a meta-skill for Claude Code or Claude Cowork
- automate skill creation instead of writing skills from scratch
- run a self-observing skill that logs its own blind spots
When to choose
- you already use an AI agent environment with a skills system (Claude Code, Claude Cowork, Codex, Hermes, Openclaw)
- you maintain a growing library of skills that never get updated
- you want systematic capture of corrections and preferences without manual journaling
- you want skill creation driven by observed work patterns rather than blank-page authoring
When to avoid
- you don't use AI agents or skill-based workflows at all
- you want fully automatic skill edits — the observer only produces recommendations for human review
- you need a standalone application rather than a skill loaded into an AI companion
- you're unwilling to review and approve observation logs regularly
Facets
plugin · maturity active
agent-framework prompt-engineering workflow-automation developer-tools artificial-intelligence developer-tools cross-platform cli meta-skill claude-skills skill-management self-improving augmented-expertise claude-code claude-cowork agentic-skills observation-logging ai-agents automation
2 sources
- readme: https://github.com/rebelytics/one-skill-to-rule-them-all · fetched 2026-09-26 · 76e8a207b96a
- homepage: https://www.rebelytics.com/task-observer/ · fetched 2026-08-29 · c62cbcb98bdf
Member repositories
| Repository | Role | Health v2 |
|---|---|---|
| rebelytics/one-skill-to-rule-them-all | main | 80 |
For agents
markdown · JSON · MCP: product_card(name="rebelytics/one-skill-to-rule-them-all")
Data as of 2026-10-19T08:05:51.212177+00:00 · Report a problem