An indie hacker is not a freelancer and not a venture-backed founder. The freelancer trades hours for invoices; the founder trades equity for runway. The indie hacker ships a product and lives off whatever MRR it generates — no clients, no board, no buffer. The bottleneck is rarely engineering. It is the surrounding loop: a landing page that converts, a price the market accepts, instrumentation that tells you what is working, an experimentation rhythm that compounds answers, a free-tool wedge, a launch that does not fizzle on day two, and a referral mechanism that turns the first 100 users into the next 1,000. The eight skills below address that loop. Each is a real, verified Claude Code skill from a plugin with public commit history and real star counts on GitHub.

From the ai-agent-camp plugin (358 stars, 3 contributors). A landing-page workflow that walks an indie hacker from a one-sentence value prop through messaging organisation, wireframe, design, and HTML implementation. Most indie products die on the landing page, not the product — the visitor lands, cannot tell in three seconds what the thing does or who it is for, and bounces. The skill enforces the structural work most solo builders skip: who is this for, what problem does it solve, why is the alternative worse, what does it cost, what is the proof. Once that is on paper, the HTML is the easy part.

When to use: on day one of a new product, before any code beyond the auth flow is written. Re-run any time the homepage conversion rate drops below a baseline you trust. Pair it with the solo-founders skill stack for the strategic groundwork that should happen before the landing page.

Also from ai-agent-camp. Supports pricing design, packaging, and monetisation — tier definition, anchor pricing, annual versus monthly, free-tier limits, and the eternal "should I add a usage component" question. Pricing is the single decision with the biggest leverage on revenue and the one most consistently made by gut feel. The skill produces three candidate tiers with rationale, anchors them against reference points the visitor recognises, and forces a decision on the free-tier boundary instead of leaving it as an accidental product of what the database happened to allow.

When to use: before the first paying customer, every time MRR plateaus for more than two months, and before any major feature ships. Re-run on a quarterly cadence — most indie hackers underprice by 30 to 50 percent and find out years late.

Also from ai-agent-camp. Sets up GA4, conversion events, UTM parameters, GTM configuration, and the event taxonomy that makes the dashboards readable six months later. The product without analytics is a product running blind — you think you know which page converts and which campaign brought the user, but you do not, and every decision after that is a guess. The skill installs the instrumentation that turns guesses into a working feedback loop, with an event naming convention that prevents the "what does click_button_2 mean" problem on month nine.

When to use: on day one, right after deploying the landing page and before running a single ad. Re-run any time you add a new conversion surface. Pair it with the data-analysts skill stack once you have months of data and need real cohort analysis.

Also from ai-agent-camp. Designs and helps implement A/B tests — hypothesis framing, variant design, sample-size estimation, and the reading-the-result step most indie hackers get wrong. It prevents two failure modes: running a test never powered to detect a realistic effect, and reading a noisy mid-experiment number as if it were final. For a one-person company with a few thousand monthly visitors, the sample-size calculator will frequently tell you the honest answer is to ship the better-looking variant and skip the test.

When to use: any time you are about to change the headline, the CTA, the pricing layout, or the onboarding step order. Run the design phase before launch, the analysis phase after, and never peek in between. Pair it with analytics-tracking above so the events the test depends on are actually captured.

Also from ai-agent-camp. Plans "engineering as marketing" — the free tool, calculator, or single-purpose utility that sits next to the main product and pulls in qualified traffic that converts over time. The pattern is the most consistent winning move in indie hacking: HubSpot did it with the website grader, Stripe with radar test cards. The skill walks through candidate tools (what does your audience repeatedly search for, what is one step in your main product that could stand alone) and scopes the smallest version that ships in a weekend.

When to use: month two, once you know who the customer is but before you have run out of organic ways to reach them. Build one free tool a quarter and let the SEO compound. Pair it with the marketers skill stack for distribution, and with analytics-tracking above so you can measure which tool drives signups.

From the gstack plugin (104,138 stars, MIT, 10 contributors — the largest plugin in the index). A structured land-and-deploy workflow that takes a feature branch from passing tests to live production with the safety rails an indie hacker wants: pre-deploy review, canary verification, rollback plan if the canary degrades. For a one-person company the production deploy is the most dangerous five minutes of the week — no on-call rotation, no SRE checking the migration, no one but you to notice when the dashboard goes blank on a Friday afternoon. The skill bakes the discipline into a workflow that survives the weeks when you would otherwise skip it.

When to use: every deploy bigger than a copy fix. Configure it with setup-deploy on day one. Pair it with the devops-engineers skill stack for the observability and incident-response half.

Also from ai-agent-camp. Plans product launches, feature releases, and GTM sequencing — phased launches, channel strategy, momentum building, and Product Hunt preparation. The difference between a launch that gets 500 upvotes and one that gets 50 is almost never the product; it is whether the founder did the four-week prep before launch day (the email list, the supporter network, the hunter, the assets, the timezone-aware launch hour, the rolling reply cadence). The skill produces the checklist, calendar, and asset list, sized for a solo operator.

When to use: four weeks before any Product Hunt launch, two weeks before a major feature release. Pair it with the content-creators skill stack for launch-content production.

Also from ai-agent-camp. Designs referral programs, affiliate programs, and word-of-mouth strategies. For an indie hacker paid acquisition is rarely viable — the LTV is too small, the CAC math does not work, and ad accounts get suspended at the worst possible moment. The growth that does work is referrals: the happy customer who tells two friends, the affiliate who lists your product in their newsletter, the social-proof loop that turns the first 100 users into a trickle of the next 1,000. The skill designs the mechanism — incentive structure, trigger moment, asset (referral code, unique URL, partner dashboard), and tracking — sized for a product without a partnership team.

When to use: once retention is stable (a referral program on a leaky bucket accelerates the leak), once the first 100 paying customers have settled in. Pair it with analytics-tracking for attribution and pricing-strategy so the incentive does not blow up the unit economics.

How to install

Seven of the eight skills live in the ai-agent-camp plugin (358 stars on minicoohei/ai-agent-camp, refreshed the week of writing) and one in gstack (104,138 stars) — a two-marketplace install. Skill Index has the exact install command on each skill detail page with a copy button. The highest-ROI sequence: install gstack first so land-and-deploy is wired up before the first production push, then add ai-agent-camp and run lp-designer and pricing-strategy before opening signups. Layer analytics-tracking in immediately so the loop has data, then ab-test-setup the first time you want to change the headline. Reach for free-tool-strategy in month two when organic reach plateaus, run launch-strategy four weeks before any moment that matters (Product Hunt, a Hacker News post you actually care about), and add referral-program once retention is stable. Pair the workflow with deep-work blocks via focus.thicket.sh and QR codes for in-person events via qr.thicket.sh, and the path from "I have an idea" to ramen-profitable MRR stops being a dice roll and starts being a sequence you can run.