Crypto Accumulation Research

Long-horizon plan · verified bounties · scam avoidance

Verified crypto earning opportunities

Checked: 3 September 2026 03:22 UTC (KeeperHub #2105 + deployer + Alchemix V3 strategies)
Scope: Legitimate, no-capital developer work that can pay crypto or stablecoins. Competitive prizes are not guaranteed income.

Ranked opportunities

Rank Opportunity Reward and deadline Effort and eligibility Risk Verdict
1 KeeperHub feature bounty $1,000 in stablecoins: two $500 winners; build window Sep 6–18 Medium–high; software engineering; 18+, solo or team, sanctions restrictions No capital risk; competition and time risk Best direct-to-crypto route. Ship a mergeable feature to the open-source repository. Do not implement before Sep 6.
2 ETHOnline 2026 Async remote hackathon Sep 4–16. Official prize page now HTTP 200 (22:56 UTC). Partner totals on that page: The Graph $15k, Hedera $15k, Arc $10k, World $7k, 1inch $7k, ENS $5k, Uniswap Foundation $5k, Ledger $5k, Privy $5k, Bazantic $3k, Chainlink $2.5k. Hedera’s largest track is still $6k x402 agentic payments ($2k × 3). High; public GitHub repo, demo video, ETHGlobal application. The Graph “Start Fresh” pool forbids project-specific prior code. Competitive. Prize rails are sponsor-specific (not a guaranteed on-chain USDC payout from this VM). Official rules are readable. Do not start project-specific code before Sep 4 (would burn Start Fresh). Still needs the participant’s ETHGlobal account and a public GitHub repo they control. See ethonline-2026-prizes.md.
3 BUIDL CTC 2026 Fall $15,000 pool: $10k/$3k/$2k; deadline Sep 13, 23:59 ET High; Attestcoin integration, testnet deploy, README, deck, video, plus each member’s legal name, email, and country. Rechecked 03:07 UTC 3 Sep: still open, 47 BUIDLs / 203 hackers, page still says “11 days left”. High execution risk; prize is USD-denominated, not confirmed as crypto CreditPassport is built in-tree; testnet deploy and DoraHacks submit still need the participant.
4 KeeperHub live-project integration $4,000 in stablecoins: $2k/$1.2k/$800; deadline Sep 18 High; must integrate with a real, already-running project and show a KeeperHub transaction High integration and judging risk Attractive only with access to a qualifying live project; a generic wrapper does not qualify.
5 Immunefi bug bounties Official explorer: “Showing all 183 bounty programs,” metrics updated 2 Sep 2026 16:01 UTC. Visible maxima include Ethena $3M and DeXe $500k Very high security expertise; program-specific KYC and proof-of-concept rules Severe legal and operational risk outside published scope Legitimate, but not quick. Stay strictly inside scope and safe-harbor terms.
6 Superteam AGENT_ALLOWED opens (Steve Arena, ZNS) $500 USDC each Steve: X + 5 mainnet trades (≥10 USDC). ZNS: launch a token and farm volume Capital and social-account risk; ZNS is a volume contest Do not execute from this VM. Agent registration is parked for a later engineering listing.
7 Superteam HUMAN_ONLY engineering-adjacent (Mermail skill, T3N agent) $500 / $290 USDC Mermail PR target is Nudgen-Marketing/mermail-skills plus a live X demo tagging @Mermailapp. Needs the participant’s GitHub, Mermail console, and X. T3N needs Terminal 3 SSO/DID and a public Google Doc. Submission-account risk only Mermail skill built 3 Sep (mermail-onchain-receipts/). T3N Vendor Receipts built 3 Sep (t3n-vendor-receipts/); remaining work is your SSO, two keys, Google Doc, and Superteam form.
8 Agent Bounties standing meta-parents on Base 2.00 USDC parent reward minus a 1.00 USDC funded child (gross ~1 USDC) plus a 0.01 USDC refundable claim bond; gas extra. Rechecked live via list_autonomous_bounties + get_bounty_feed at 01:08 UTC 3 Sep: 7 claimable routed parents (API reliability, MCP, CLI, wallet UX, discovery, distribution). Needs a signable Base wallet, ~1.02 USDC + gas, and a different registered participant to settle the child (self_verification_forbidden). Readiness: POST https://api.agentbounties.app/v1/base/agent-wallet/readiness. Legal if the child is real work a stranger completes. Dual-wallet farming and GMV “highest volume” leaderboards are not an acquisition strategy. Next funded-wallet job, not executable from this VM today. Syndicated “Binance Agent OS $60k Mini Hackathon” news had no official binance.com submission page — do not build against it.

Dead ends checked

KeeperHub candidate queue

Repository state changes quickly. Recheck issues and pull requests before claiming anything.

Candidate Evidence on Sep 2 Decision
#2105: OpenAPI workflow-call response examples Rechecked 03:12 UTC 3 Sep: issue still open + accepted + confirmed. GitHub search repo:KeeperHub/keeperhub 2105 type:pr returned 0. Last issue update 2026-08-20. Patch spec in keeperhub/2105-openapi-response-examples.md. Primary candidate after Sep 6. Do not implement against upstream before the build window.
#2097: protocol-action preflight and idempotency Accepted and confirmed; no linked or open PR found. High safety value, but check overlap with #2004 and #2207 first.
#2062: onboarding ID glossary and cross-links Accepted with maintainer-narrowed scope; no linked PR found. Low-risk documentation fallback; weaker “feature” fit.
#2247: trace-method provider survey Accepted, good-first-issue, help-wanted, apparently unclaimed. Useful research, but the bounty asks for a feature.
#2240: threshold-over-state trigger Accepted/help-wanted; unresolved design questions cover edge detection, re-arming, deduplication, and scaling. Strong product value but too broad for quick work without maintainer decisions.

Already have PRs — do not duplicate: #2208 (PR #2215), #2211 (PR #2217), #2206 (PR #2213), #2230 (PR #2228), #2196 (PR #2197). GitHub timeline check 2 Sep 2026.

Contribution gate for candidate #2105

KeeperHub's contribution policy requires:

The current source confirms the issue precisely: app/api/openapi/route.ts has separate hardcoded read and write success schemas with no whole-response examples, while tests/unit/openapi-route.test.ts is the focused verification surface. No dependency, database, wallet, or mainnet transaction is required.

The clone's default branch is staging, and its HEAD matched the freshly fetched origin/staging commit d249519 during this check. GitHub search returned zero open pull requests referencing 2105. OpenAPI 3.1 uses a bare examples array inside a Schema Object, while a Media Type Object uses either a singular example value or a named examples map. Because #2105 specifically asks for values on both hardcoded response schemas and KeeperHub already uses schema-level examples arrays, schema-level arrays are the most consistent implementation candidate. Recheck all of this after Sep 6.

Build feasibility was verified without changing KeeperHub source: the required Node.js 24 runtime and pnpm 10 were activated, the frozen lockfile installed successfully, and pnpm exec vitest run tests/unit/openapi-route.test.ts passed all 27 tests on the Sep 2 staging commit. pnpm check also passed with one pre-existing oversized-fixture warning. A fresh-clone pnpm type-check initially failed because generated registries were absent; the documented pnpm discover-plugins step generated them, after which type-check passed with no tracked source changes. This is baseline evidence only; it is not submission code and does not satisfy the bounty.

Resources and blockers

Needed from the participant:

Potentially needed later:

Not needed: investment capital, trading deposits, exchange API keys, seed phrases, or private keys.

Execution order

  1. Confirm participant eligibility.
  2. Recheck #2105 and the open pull-request list when the build window opens Sep 6.
  3. Comment to claim the issue before coding so contributors do not duplicate work.
  4. Implement with a failing test first, then run the focused suite and required repository checks.
  5. Open a pull request against staging.
  6. Register and submit a separate BUIDL for the KeeperHub feature bounty with the source link and required evidence.
  7. Treat an accepted PR and a submitted BUIDL as progress only. Earnings are verified only after an official winner result and an onchain payout to the participant's own wallet.