The contributor flywheel is live: ladder, invitations, and how paid work routes

Two weeks ago we wrote the thesis: a research collective is a community problem before it is a technical one. Today we shipped the machinery. This post is a plain description of what is now running in production, who it is for, and how the money routes.

What shipped

Three small pieces of infrastructure, all open-source in this repo:

• Contributors ledger — rebuilt daily from the GitHub API across every kcolbchain repo. On first run it found 16 external contributors with at least one merged PR. It ships tier badges (Newcomer / Regular / Fellow), repos touched, and a summary count. Opt-out is one line in data/contributors-optout.txt.
• Open invitations board — every open issue across the org tagged good first issue, help wanted, or bounty. 50 invitations on day one. Filter by label, click through to the issue. Policy-safe: a label on its own is not an offer of payment; any reward must be stated inside the issue body.
• Post-merge thank-you — when the PR pipeline merges your PR, you get a comment. First-merge contributors get the longer version: what unlocks next, where the ladder goes, what the Fellow tier means.

The ladder (signal, not contract)

We use three public tiers derived from merged-PR counts across the org. They are visibility signals — not contracts, not entitlements. Full text lives in §4a of the contribution policy.

At the time of writing: 3 Fellows, 5 Regulars, 8 Newcomers. The bar is lower than most bounty programs on purpose — five small, landed PRs tell us more about fit than one heroic drive-by ever did.

Where the money actually comes from

This is the part most OSS community posts skip, so here is ours in plain language. kcolbchain runs on a mix of four revenue streams:

1. Partner retainers. We are currently engaged with Liquidity.io on a fixed monthly retainer for work scoped to their GitHub org. Similar structures exist with other partner teams; scope and terms are per engagement.
2. Grants. Ecosystem programs (Optimism RPGF, Ethereum Foundation, Arbitrum, Uniswap, and others) fund specific deliverables with defined milestones. When we win one, the work gets sliced into issues with stated terms.
3. Applied consulting. Protocol teams hire us for short scoped engagements: audits, research briefs, launch-risk reviews, architecture second opinions.
4. Co-published research. Long-form notes co-authored with protocols we collaborate with. Sometimes sponsored, sometimes paid via the partner's research budget.

When any of those streams produces work that a Fellow is a good fit for, we invite them first. That is the entire mechanism. It is not a bounty program because we are not advertising a fixed price for pre-defined tasks. It is closer to preferred subcontracting: we have a small circle of people we trust to ship, and when paid work that matches their areas shows up, they get the first message.

Crucially: Fellows are not obligated to say yes, and kcolbchain is not obligated to offer. Both sides stay optional. What the tier actually guarantees is first to hear, not first to be paid.

Why this structure

The open-source world has three roughly working models for keeping maintainers fed:

• Corporate backing (Linux Foundation, Node.js style): a company pays a salary; the person works on the project. Stable, but narrow — only a few get in.
• Grants (Ethereum Foundation, NumFOCUS): lump sums for defined deliverables. Works but cycle-dependent, and grant fatigue is real.
• Bounties (Gitcoin, Immunefi): pay-per-task. Scales, but distorts incentive toward the easiest-to-scope tasks, not the most important ones.

The ladder + invitations + contributor Telegram model is a fourth shape — a relationship-first pipeline. Nobody signs anything to contribute. Nobody is promised anything by picking up an issue. But contributors accumulate trust through merged PRs the same way a junior dev accumulates it on a team, and the work we route to them scales up as that trust grows. It is less efficient than a well-priced bounty for narrow tasks, and more efficient than pure drive-by OSS for anything requiring context.

It also maps better onto how paid work actually reaches us. When a partner says "we need someone who knows x402 and can ship in Rust by Friday," we do not post a bounty. We DM a Fellow. That is already how the world works — we are just making the ladder visible so everyone knows the rules.

What the frontier looks like right now

If you want to pick up where the work actually is:

• invitations/ — 50 open issues across 14+ repos, filtered by good first issue, help wanted, and bounty. Pick one. The PR pipeline auto-triages, auto-audits (Claude Haiku), and auto-merges anything clean within a couple of hours.
• ROADMAP.md — our Q2 commitments. The items that still have open boxes are where new contribution makes the biggest dent.
• docs.kcolbchain.com — the user-facing side of every shipping project. Pick the one that looks interesting and find its repo from there.

What still needs to ship

A partially built flywheel is worse than useless; it looks like motion without producing force. Two pieces are not here yet, and we will say so out loud:

1. The contributor Telegram is not open yet. We reference it throughout the ladder docs. The channel exists; the invite flow does not. We will ship that this week.
2. There is no public log of invited paid work. Fellows currently hear about paid opportunities via DM. That is fine for now but opaque. We want a private page — contributor-auth gated — that lists opportunities as they come in. Probably in Q3 once we have more flow.

If you made it this far and you are a blockchain engineer, a protocol researcher, or just a curious developer who likes shipping over philosophy: the fastest path in is one merged PR. We will take it from there.

Start here: browse open invitations · see who is building · read the policy