Universal Planning Framework

Universal Planning Framework

https://github.com/user-attachments/assets/6358ef3b-24e5-4b19-9c0e-bcce22a234ee

Plans fail because discovery happens too late. This framework catches gaps that only surface during execution - evolved from 117 real plans + 195 handoffs.
"Initial idea was a custom booking system (8 weeks). Stage 0 discovered Calendly + Stripe does 90% of it. Shipped in 3 weeks, saved 5 weeks of engineering."
- from Business Launch Example

Quick Install

As a plugin (recommended - nothing is copied into your project):

Run both inside Claude Code. You get the four commands namespaced, so
/plan-new becomes /universal-planning-framework:plan-new, plus the planner
agent and two skills Claude reaches for on its own, universal-planning and
vehicle-selection, which carry the rulebook and the vehicle rubric.
Update later with /plugin marketplace update primeline-upf.

Into your project's .claude/ (plain files you can read and edit):

setup.sh never replaces a file you already have. If one of its file names clashes
with yours it asks, and skips by default. --dry-run shows what it would do,
--skip-existing answers every clash without asking, and --overwrite replaces
after writing a timestamped backup next to each file. Commands land as /plan-new
and friends, with no namespace prefix.

Minimal (the rulebook alone, no commands):

What Makes This Different

Three things carry most of the value.

Discovery happens before the plan, not during execution. Stage 0 runs 12 checks
before a single phase is written, and the one that pays for the rest is the AHA check:
*does something that already exists do most of this?* A custom CMS becomes Strapi at
90% of what you needed. Fifty blog posts become five deep ones plus derivatives. The
cheapest moment to find that out is before…