The Honest Negative Result

This final paper walks the full lifecycle of winning a real open-source bounty — from discovery through claim, fix, review, and payout — and reports what happened when we tried to execute it live. The headline is a negative result, reported honestly: the naive "pick a task off a popular board and win it fast" thesis (H4 from Part 1) was falsified by live data. Consistent with GitScout's anti-fabrication premise, we did not manufacture a completed claim.

01The idealized lifecycle

The end-to-end path a bounty should follow:

1. Discover a real, funded, unassigned bounty on your stack.
2. Read the rules. On Algora: comment /attempt #N with an implementation plan; don't ask to be assigned unless you've contributed before; UI changes require a demo video; "low-quality AI PRs will not receive review and will be closed."
3. Research & reproduce the issue against the real codebase.
4. Fix it with a clean, reviewable change.
5. Propose a PR whose body contains /claim #N.
6. Review loop with the maintainer until merge.
7. Get paid — "100% of the bounty is received 2–5 days post-reward", i.e. USD via Stripe Connect.

Steps 1–2 are where most of the value — and most of the failure — actually lives.

02What happened live (verified 2026-09-03)

We executed steps 1–2 for real, deep-diving four candidates in the live Algora and GitHub UIs:

The whole tscircuit org board: 10 open / 707 completed. The remaining open items are the picked-over remainder — old, tiny ($1–$170), and heavily claimed. Opire, the less-crowded alternative we wanted to fall back on, was returning HTTP 502 at check time.

03The negative result, stated plainly

Two temptations were available here, and both were refused:

1. Attempt a saturated issue anyway to produce a "win" narrative. Rejected: expected value is ~$0, and a rushed AI PR into a repo that explicitly closes low-quality AI PRs would damage the user's real reputation.
2. Write up a claim that never happened. Rejected outright — fabricating a completed bounty is exactly the fraud GitScout exists to detect. The credibility of an anti-fraud tool cannot survive its author faking a result.

Reporting the negative result is the deliverable here.

04Why this strengthens the product

The failure exposed a second axis of worthlessness that a raw feed cannot see. A bounty wastes your time if it is either not real (farm/scam — handled by the Part 2 trust engine) or real but unwinnable/already-won (saturated, stale, or already rewarded). GitScout already scores the first. The live data proves it must also score the second — a viability/liveness signal:

• Attempt density — count of /attempt claims; >~10 ⇒ red.
• Already-rewarded flag — presence of a "Rewarded" label / award-bot comment ⇒ hard-hide from the "open" view.
• Veteran lock-in — a hunter with a large completed count already attempting.
• Staleness — a large, old, still-open bounty is usually a graveyard.
• Newcomer-eligibility — boards that gate assignment on prior contribution.

05The realistic strategy that actually wins

1. Reset expectations. A first win is likely $3–$50 and is reputation-building, not income — the price of unlocking the "contributed before" gate.
2. Compete on freshness, not on picked-over issues. The winnable bounty is the one posted minutes ago with zero attempts. This motivates GitScout's next feature: a new-bounty monitor that alerts on freshly-funded issues on your stack so you can be the first credible /attempt.
3. Build reputation with non-bounty PRs first — one or two small, clean merged PRs to become a known contributor before racing for cash.
4. Diversify platforms — re-check Opire when it's up (less crowded, ~4% fee); consider Polar-funded issues; avoid crypto-native platforms unless you specifically want them.
5. Verify at the issue, every time — amount, not already rewarded, attempt count, and whether a veteran is on it — before writing any code.

06Getting paid — the "wallet" question, corrected

A natural request is to "create a wallet to receive a claimed bounty." Two corrections:

1. The realistic (Algora/Opire) path needs no crypto wallet. Payouts are USD via Stripe Connect / bank transfer, 2–5 days post-reward. A crypto wallet only enters via (a) the bounty-plaza scam we filter out, or (b) optional crypto-native platforms (Gitcoin/Immunefi).
2. The assistant cannot and did not create an account, wallet, or payout instrument. Creating accounts and entering financial/identity details are the user's own actions. The concrete steps, for the user to perform: sign in to Algora (or Opire) with the existing GitHub account; complete Stripe Connect payout onboarding (identity + bank/debit) to become "eligible for payouts." That is the entire "wallet" for this path — no crypto.

07Conclusion of the series

GitScout set out to be a bounty aggregator you can trust. Across three papers it became something sharper: a tool that scores both authenticity (is it a scam?) and viability (can you actually win it?), and whose own development honored the same standard by reporting a falsified hypothesis instead of hiding it. The next build is the freshness monitor and the viability signal; the next action for the user is Stripe payout onboarding and a reputation-first first contribution — not a race into a saturated board.

All bounty amounts, states, and attempt counts were read from the live Algora / GitHub UI on 2026-09-03. No bounty claim was submitted, and none is represented as having occurred.