v1.2.1 — OpenAI category compatibility

Patch release for OpenAI plugin submission compatibility.

Fixed

• Replace the unsupported Lifestyle plugin category with the allowed Entertainment category.
• Add the OpenAI plugin category allowlist to CI so unsupported categories fail before submission.
• Keep the Claude Code, Codex, portal-copy, and archive versions aligned at 1.2.1.

Use hinge-profile-optimizer-1.2.1.zip for the OpenAI Skills only submission.

SHA-256: c1c7e6f9daa7ccea6c2b3f04200ac128e4a83a37f0cbf3fe229cd3fe4d72b99b

v1.2.0 — ChatGPT and Codex plugin

Hinge Profile Optimizer is now packaged as a skills-only plugin for ChatGPT and Codex while preserving the existing Claude Code and standalone skill installs.

What's new

• Add Codex/OpenAI and Claude Code plugin manifests.
• Add a self-contained packaged skill with all seven research references.
• Add ChatGPT/Codex interface metadata, starter prompts, and square plugin artwork.
• Add privacy and terms pages plus OpenAI submission copy and behavioural test cases.
• Add explicit photo-review, identity-inference, honesty, and consent boundaries.
• Add reproducible ZIP builds, drift detection, and GitHub Actions validation.
• Refresh the README and repository About metadata for ChatGPT, Codex, and Claude Code.

OpenAI submission

Download hinge-profile-optimizer-1.2.0.zip from this release and upload it as a Skills only plugin. The prepared portal copy and test cases are under submission/.

The plugin has no backend, authentication, analytics, telemetry, or independent data store.

Verification

• Plugin bundle check passed on the merged main commit.
• Root and packaged Agent Skills validated.
• Codex plugin manifest validated.
• GitHub Actions passed.

SHA-256: ebf7dffdde18b14963c4bbf579c5edc8b0ce54d69b6ce23a69fe06a16d5a9a25

v1.1.0 — filtering scales to market thickness

First tagged release, and the one substantive change since the skill went public.

Filtering now scales to market thickness.

Phase 3 always estimated the viable pool and then dropped the number on the floor. Everything after it assumed the pool was large. That assumption is load-bearing for the skill's most distinctive advice: use niche references, hold strong opinions, let the wrong people bounce. It's correct when the pool is several hundred people. It's actively harmful when the pool is forty.

Phase 3 now names a thickness band — thick, moderate, or thin — and the copy principles read from it. Niche references and strong opinions soften as the pool narrows. Specificity doesn't move at any band, because concrete detail signals honesty regardless of how many people are reading it. What scales is how much alienation the market can absorb.

Thickness is deliberately not a location setting. It's metro size, age band, orientation, and dealbreaker width multiplied together — a 52-year-old in a big city and a 28-year-old in a small one can land in the same band, and two people in the same postcode often don't. A city field would have captured a fraction of the variance while implying local knowledge the skill doesn't have. The band comes out of the discovery interview instead.

One practical addition: where the pool is thin, the first move is settings rather than copy. Widening radius or age range by a notch usually beats any rewrite.

Raised by @KevinColemanInc.