This skill helps you audit a Hinge profile, find better photos, interview you for specific material, and rewrite prompts in your own voice. It uses a phased process with research notes and reference files instead of generic dating advice.
Builders who want their agent to rebuild a Hinge profile from screenshots, photos, or a fresh start.
You can turn a generic profile into one that shows your actual personality and improves match quality.
Scores current profile screenshots against audit criteria so you can see what is weak or missing.
Asks questions to pull out specific details, humor, interests, and relationship style that can be used in prompts.
Reviews candidate photos, orders them, and points out gaps using the photo guidelines.
Writes Hinge prompts and captions from your own material instead of templates.
Reviews Hinge settings so you can reduce clutter and improve visibility.
Bakes in cited findings, copy principles, and discovery questions from the reference docs.
Ships plugin manifests for Claude Code and Codex/OpenAI so the skill can be installed as a bundle.
npx skills add b1rdmania/hinge-profile-optimizer --full-depth
claude --plugin-dir ./hinge-profile-optimizer
A research-backed Agent Skill for rebuilding a Hinge profile in about 45 minutes
Everyone has something. The way they think, what they care about, their weird specific interests, how they show up for people, what makes them laugh.
Most dating profiles bury this under generic prompts and bad photo choices.
This skill finds it and puts it where people can see it.
It can audit profile screenshots, choose and order photos, interview you for better material, write prompts and captions in your voice, review settings, or diagnose weak match quality. Use the full process or ask for one part.
Independent, unofficial project. Not affiliated with or endorsed by Hinge or Match Group.
45 minutes.
Not a quick fix. Not "5 tips for better prompts."
This is the full process: honest audit, proper interview, photo strategy, copy that sounds like you, settings cleanup, and help putting it live.
Whether you want a life partner, a good date, or something casual, the profile should still be recognisably yours.
Most "profile tips" give you a template and send you on your way. Generic prompts, generic results.
This is a structured 8-phase process that actually gets to know you first:
| Phase | What Happens |
|---|---|
| Setup | Frame the process, understand your situation |
| Audit | Score your current profile (skip if starting fresh) |
| Discovery | The big interview - find what actually makes you you |
| Reality Check | Honest market math - who are you competing for? |
| Photos | Evaluate, order, identify gaps |
| Copy | Write prompts using your material, not templates |
| Settings | Optimize visibility, hide the clutter |
| Implementation | Put it live together |
| Algorithm | What to do in weeks 1-4 |
Why it works: The discovery phase is the key. Most advice is generic because it doesn't know you. This spends real time understanding your humor, opinions, relationships, weird rituals — then uses those as ingredients. Your 92-year-old great uncle who dominates pub quiz? That's going in the profile.
This isn't vibes. It's grounded in published research from actual journals, data from dating platforms, and practical experience — and we're honest about which is which.
People decide fast, and mostly from photos. Users spend seconds on a profile, and most of that time goes to photos (Brand et al., 2012). Photos drive the swipe. But on prompt-based apps like Hinge, text matters more than people assume — creative, original profiles are independently rated as more attractive, even controlling for photos (Fiore et al., 2008).
Specific language beats generic language. This is one of the strongest findings. "Jazz Cafe on a weeknight" works better than "live music" — and we know why. Toma & Hancock (2012) found that specific, concrete language is a linguistic marker of honesty. Vague, abstract phrasing is a marker of deception. Readers pick up on this without knowing they're doing it. Separately, specific language creates a feeling of closeness — naming a real place puts someone there, while a category keeps them at arm's length (Trope & Liberman, 2010).
Showing beats telling. Writing "I have dark humor" tells people nothing — anyone can claim any trait. Writing "Being nice about Timothee Chalamet" demonstrates it. Signaling theory (Donath, 2007) explains the difference: a demonstrated quality is a costly signal — hard to fake, therefore credible. A stated quality is a cheap signal — easy to fake, therefore ignored.
Humor works, but only if you're actually funny. Humor in profiles increases perceived attractiveness and reads as a signal of intelligence (McGee & Shevlin, 2009). Producing humor matters more than claiming to appreciate it (Bressler & Balshine, 2006). Failed humor is worse than no humor.
The market is unequal. Bruch & Newman (2018), published in Science Advances, analyzed 200,000 users and found that desirability follows a power law. Most people pursue partners roughly 25% more desirable than themselves. The top profiles receive wildly disproportionate attention. This is why the skill focuses on differentiation and filtering rather than broad appeal — you're not trying to be liked by everyone.
29 cited studies, platform data from Hinge and OkCupid, plus foundational work on signaling theory and self-disclosure. Every claim tagged with its evidence tier. Full citations in references/research-findings.md.
The repository keeps SKILL.md at the root for existing standalone installations and packages a self-contained copy under skills/hinge-profile-optimizer/ for plugin hosts.
Install the standalone skill directly from GitHub:
npx skills add b1rdmania/hinge-profile-optimizer --full-depth
Invoke $hinge-profile-optimizer, or ask Codex to audit or rebuild a Hinge profile.
The repository includes .codex-plugin/plugin.json for the Plugins Directory shared by ChatGPT and Codex. Once OpenAI approves and publishes the listing, install Hinge Profile Optimizer from the directory and start with screenshots, a photo set, or a request for a fresh profile.
Until publication, the ZIP attached to the latest GitHub release can be uploaded through the OpenAI plugin submission portal by the verified publisher.
Load it as a plugin from a local checkout:
claude --plugin-dir ./hinge-profile-optimizer
Existing personal-skill installations that point at the repository root continue to work.
Download the repository and add SKILL.md plus the seven files under references/ to project knowledge or the current conversation. Then ask: “Help me optimise my Hinge profile.”
The standalone skill is also listed on SkillsMP.
.claude-plugin/plugin.json exposes the packaged skill while the existing marketplace entry remains available..codex-plugin/plugin.json exposes the same skill through a skills-only plugin.Build and check the submission archive with:
bash scripts/check-plugin-bundle.sh
You're not making someone more appealing. You're making them visible.
The interesting stuff is already there. A 92-year-old great uncle who's a pub quiz champion. An irrational hatred of a specific celebrity. A hate-watch ritual every Saturday morning. The exhibition catalogue they're prouder of than anything else they've made.
These details get lost in "love to laugh" and "partner in crime."
This process pulls them out.
There's someone for everyone. They just can't find each other when every profile says the same thing.
This skill helps people show who they actually are to the people who'd appreciate that person.
MIT
Sign in to join the discussion.
No comments yet. Be the first to say what this is good for.
Open-source AI job search: scan job portals, evaluate listings into a structured A-H report with a global 1-5 score, tailor your CV, track applications — runs locally in your AI coding CLI (Claude Code, Codex, OpenCode, Antigravity…)
A skill to stop your coding agent from burying the answer. ADHD-friendly output.
AI agent skill that researches any topic across Reddit, X, YouTube, HN, Polymarket, and the web - then synthesizes a grounded summary
De-AI writing skill for any Agent Skills-compatible agent (77+ via the Skills CLI), with native plugins for Claude Code, Codex, Grok Build, and Antigravity. Narrative-architecture repair for fiction, venue-matched rules for professional prose. Based on StoryScope (arXiv:2604.03136).
Academic Research Skills for Claude Code: research → write → review → revise → finalize
Agent skill that removes signs of AI-generated writing from text