Problem
The homepage compresses UF to "three commands — Specify, Unleash, Finale" and calls it "the complete developer loop." This is memorable but hides UF's strongest differentiator: the post-PR lifecycle.
What's missing from the homepage story
-
/uf.review-pr (1,158 lines) — token-efficient PR reviewer with CI causality classification (PR-caused vs. pre-existing failures), convention-pack enforcement, compound-severity calibration, suggestion blocks, and human-gated GitHub review posting. No competitor has this.
-
/uf.address-feedback (627 lines) — four-phase feedback processor (Ingest → Assess → Triage → Execute) that classifies reviewer comments as data-driven vs. subjective, escalates to specialist Divisor personas (security → Adversary, arch → Architect), implements accepted changes with conventional commits, and posts reply comments with confirmation gates. UF-only capability; no equivalent exists in any competitor.
-
/uf.triage-issue — multi-agent issue triage with 5-agent Divisor panel, 7 classification categories, and human-gated label mutations.
The real loop
The differentiated developer loop is closer to 5-6 commands:
specify → unleash → finale → review-pr → address-feedback
↑ ↓
└──── iterate ──┘
Plus /uf.triage-issue as the front door.
Proposed
-
Extend "How It Works" section to show the post-PR lifecycle as a 4th or 5th card, or reframe the 3-card layout to acknowledge what happens after the PR opens.
-
Elevate the harness thesis (built-in review/quality/memory/triage) above or alongside the generic feature cards in "Why Unbound Force?"
-
Reconcile "three commands" messaging — it can stay as the core loop, but should not claim to be the complete loop.
Context
The post-PR lifecycle was identified as the widest capability gap between UF and competitors during the UF-vs-SPOG comparison analysis. /uf.review-pr and /uf.address-feedback have no functional equivalent in any reviewed system.
Problem
The homepage compresses UF to "three commands — Specify, Unleash, Finale" and calls it "the complete developer loop." This is memorable but hides UF's strongest differentiator: the post-PR lifecycle.
What's missing from the homepage story
/uf.review-pr(1,158 lines) — token-efficient PR reviewer with CI causality classification (PR-caused vs. pre-existing failures), convention-pack enforcement, compound-severity calibration, suggestion blocks, and human-gated GitHub review posting. No competitor has this./uf.address-feedback(627 lines) — four-phase feedback processor (Ingest → Assess → Triage → Execute) that classifies reviewer comments as data-driven vs. subjective, escalates to specialist Divisor personas (security → Adversary, arch → Architect), implements accepted changes with conventional commits, and posts reply comments with confirmation gates. UF-only capability; no equivalent exists in any competitor./uf.triage-issue— multi-agent issue triage with 5-agent Divisor panel, 7 classification categories, and human-gated label mutations.The real loop
The differentiated developer loop is closer to 5-6 commands:
Plus
/uf.triage-issueas the front door.Proposed
Extend "How It Works" section to show the post-PR lifecycle as a 4th or 5th card, or reframe the 3-card layout to acknowledge what happens after the PR opens.
Elevate the harness thesis (built-in review/quality/memory/triage) above or alongside the generic feature cards in "Why Unbound Force?"
Reconcile "three commands" messaging — it can stay as the core loop, but should not claim to be the complete loop.
Context
The post-PR lifecycle was identified as the widest capability gap between UF and competitors during the UF-vs-SPOG comparison analysis.
/uf.review-prand/uf.address-feedbackhave no functional equivalent in any reviewed system.