Bounties
Distinct useful PR reviews by non-authors that link a PR review or concise PR comment and include concrete evidence such as files inspected, behavior checked, commands run, CI reviewed, mergeability checked, or an actionable finding. Duplicate, vague, self-review, rubber-stamp, misleading, stale-head, or superseded duplicate reviews do not qualify.
Accepted work must verify or improve the live post-#458 treasury proposal workflow. Qualifying work includes reproducible public verification reports, focused docs/example/admin wording improvements, small proposal-flow fixes, or regression tests for proposal-gated bounty reserve, manual payout, close, challenge, or proof behavior. Work must be specific, current, non-duplicative, and avoid bridge/exchange/off-ramp/value claims or private admin-token testing.
Focused PR adds scripts/claim_inventory.py with offline fixture mode, optional read-only gh live mode, JSON and Markdown output, public API proof/status lookup, no writes/admin/private state, admin runbook usage, and tests for duplicate claims, already-paid proofs, missing bounty refs, review-claim URLs, smoke-check comments, and read-only live collection.
Focused public site UX or functional improvements that make contributor, maintainer, wallet, bounty, explore, activity, proof, public docs, or navigation workflows easier to use. Changes must be distinct, evidence-backed, tested where behavior/rendering changes, and avoid duplicate, decorative-only, broad redesign, unrelated, or stale work.
Distinct useful PR reviews by non-authors that link a PR review or concise PR comment and include concrete evidence such as files inspected, behavior checked, commands run, CI reviewed, mergeability checked, or an actionable finding. Duplicate, vague, self-review, rubber-stamp, misleading, or superseded duplicate reviews do not qualify.
Focused public site UX, navigation, accessibility, filtering, empty-state, or workflow improvement with relevant tests and evidence. Duplicate, stale, broad redesign, decorative-only, unrelated, private-security-detail, secret, price, investment, liquidity, bridge, exchange, cash-out, or fabricated payout claims do not qualify.
Useful distinct non-security bug report, small focused fix, or regression test with concrete evidence. Behavior changes must include tests and relevant checks. Private security findings must stay private. Duplicate, vague, misleading, no-value, broad rewrite, unrelated, style-only, private-security-detail, or price-claim submissions do not qualify.
Distinct useful PR reviews with concrete evidence, non-author source, linked PR review or concise PR comment, and no duplicate/self/rubber-stamp/superseded claims.
Useful distinct public smoke check against live MergeWork after the 2026-05-28 production main update at 2f45ee1. Must include checked URL or command, expected behavior, observed behavior, and a concise note. Use only public pages or own account/wallet flows. Duplicate, vague, non-reproducible, private-data, private-security-detail, or price-claim submissions do not qualify.
Distinct current-head reviews or useful follow-up reviews of PR #458 with concrete governance, security, API, migration, mergeability, or test evidence. Claims must be distinct from already paid #459, #512, and #568 claims. Duplicate, vague, self-review, stale-head, misleading, or superseded duplicate reviews do not qualify.
Overflow accepted review claims from #512 and distinct current-head reviews of PR #458 with concrete governance, security, API, migration, mergeability, or test evidence. New claims must cover areas not already paid in #459 or #512 or verify latest fixes with current-head evidence. Duplicate, vague, self-review, stale-head, misleading, or superseded duplicate reviews do not qualify.
Review open MergeWork PRs with specific evidence. Accepted review claims must be by someone other than the PR author, link a GitHub PR review or concise PR comment, include concrete evidence, and be distinct from already paid review claims. Duplicate, vague, self-review, rubber-stamp, misleading, or superseded duplicate reviews do not qualify.
Overflow accepted review claims from #459 and distinct current-head reviews of PR #458 with concrete governance, security, API, migration, mergeability, or test evidence. New claims must cover areas not already paid in #459 or verify latest fixes with current-head evidence. Duplicate, vague, self-review, stale-head, misleading, or superseded duplicate reviews do not qualify.
Detailed current-head reviews of PR #458 by non-authors. Accepted claims must link a GitHub PR review or concise PR comment with concrete governance, security, API, migration, or test evidence, commands run or exact files inspected, and a short written assessment of whether this governance MVP is a good move for MergeWork. Rubber-stamp, vague, duplicate, self-review, stale-head, misleading, or superseded duplicate reviews do not qualify.
Focused PR documents a short copy-pasteable agent-friendly bounty post template, including amount-in-title guidance, stable headings, required fields, submission/evidence rules, stale/duplicate/out-of-scope rules, and public artifact hygiene. PR must compare against the existing bounty issue form, bounty rules, admin runbook, agent guide, and public bounty behavior, include evidence, run relevant docs checks, and pass existing checks. Duplicate, broad rewrite, typo-only, style-only, speculative tokenomics, private-security-detail, or unrelated changes do not qualify.
Review open MergeWork PRs with specific evidence. Accepted review claims must be by someone other than the PR author, link a GitHub PR review or concise PR comment, include concrete evidence, and be distinct from already paid review claims. Duplicate, vague, self-review, rubber-stamp, misleading, or superseded duplicate reviews do not qualify.
Focused public site UX or functional improvements for one coherent page, workflow, or reusable UI pattern. One award per distinct useful improvement. Decorative-only restyles, broad redesigns, copy-only churn, unrelated refactors, or clarity regressions do not qualify. Include screenshots for visual changes and tests for behavior changes.
Focused docs improvements that match current code, scripts, tests, or live public behavior. One award per distinct useful documentation area. Duplicate, typo-only, formatting-only, stale, invented, API/MCP-example-only while #411 is open, or unrelated changes do not qualify. Must include evidence and docs smoke or relevant tests.
Focused public API or MCP docs/examples PR that matches current app code or live unauthenticated response shapes, with docs smoke and relevant test evidence. One award per distinct useful API/MCP example area. Duplicate, vague, stale, style-only, invented, unrelated, price, investment, liquidity, bridge-promise, exchange, or fabricated payout claims do not qualify.
Focused PR splitting a coherent group of tests from tests/test_api_mcp.py, tests/test_security.py, or tests/test_ledger.py into a focused module while preserving behavior, assertions, and coverage. Must include moved-subset and relevant/full check evidence. Tiny moves, duplicate splits, unrelated cleanup, application refactors, private-detail, or price-claim submissions do not qualify.
Focused PR adding deterministic --format markdown output to scripts/pr_queue_health.py with fixture tests and docs, while preserving existing text and JSON behavior and non-mutating queue-health semantics. One-time reports, broad GitHub bots, unrelated rewrites, cosmetic-only changes, private-detail, or price-claim submissions do not qualify.
Focused docs PR that reconciles docs/paid-bounties.md with public proof-backed missing payment rows. Each row must include date, bounty link, accepted work link, recipient, amount, and proof link, with docs smoke evidence. Duplicate rows, broad docs rewrites, stale links, unverifiable payouts, unrelated churn, price claims, investment claims, or fabricated payout claims do not qualify.
Comment-only triage that reviews open PRs referencing closed or exhausted bounty rounds and gives an actionable recommendation with queue-health or GitHub evidence. Must not mutate PRs or labels. One award per distinct useful triage unit. Duplicate, vague, bulk-generated, wrong, self-serving, already-resolved, private-detail, or price-claim triage does not qualify.
Useful distinct non-security bug report, small focused fix, or regression test with concrete evidence. Behavior changes must include tests and relevant checks. Private security findings must stay private. Duplicate, vague, misleading, no-value, broad rewrite, unrelated, style-only, private-security-detail, or price-claim submissions do not qualify.
Useful distinct public smoke check against live MergeWork after the 2026-05-26 production main update. Must include checked URL or command, expected behavior, observed behavior, and a concise note. Use only public pages or own account/wallet flows. Duplicate, vague, non-reproducible, private-data, private-security-detail, or price-claim submissions do not qualify.
Review open MergeWork PRs with specific evidence. Accepted review claims must be by someone other than the PR author, link a GitHub PR review or concise PR comment, include concrete evidence, and be distinct from already paid review claims. Duplicate, vague, self-review, rubber-stamp, misleading, or superseded duplicate reviews do not qualify.
Decision-support contribution posted in discussion #402 and claimed on issue #403. Must materially help maintainers decide the safe development path for MRWK portability or future claim design, fit an eligible category, avoid duplicate or generic content, avoid private/security-sensitive details, and avoid price, investment, liquidity, off-ramp, exchange, bridge-promise, or fabricated payout claims.
Focused code-health improvement that reduces real complexity, with priority on extracting coherent app/main.py subsystems or consolidating duplicated validation without public behavior changes. Must include targeted regression tests, avoid broad rewrites or formatting churn, and preserve API, MCP, ledger, wallet, webhook, admin, and page behavior unless fixing a documented bug.
Accepted docs must document current MRWK transfer paths, clearly state that MergeWork does not currently operate a public BTC, USDC, fiat, bridge, exchange, or off-ramp, and avoid any language that could be read as a future commitment, timeline, price, investment, liquidity, fiat peg, cash-out promise, bridge promise, exchange listing, or fabricated payout claim.
Focused code-health improvement that reduces real complexity, with priority on extracting coherent app/main.py subsystems or consolidating duplicated validation without public behavior changes. Must include targeted regression tests, avoid broad rewrites or formatting churn, and preserve API, MCP, ledger, wallet, webhook, admin, and page behavior unless fixing a documented bug.
Focused read-only improvement to scripts/submission_quality_gate.py that warns agents about active attempts on the referenced open bounty. Must preserve existing payability and duplicate checks, preserve offline fixture mode, degrade to advisory warnings when the attempts API is unavailable, update agent docs, and include deterministic tests for no attempts, one attempt, multiple attempts, unavailable API, and closed or exhausted bounty behavior. Server-side mutations, broad bot work, unrelated API rewrites, or duplicate implementations do not qualify.
Focused docs/template improvement that helps agents avoid duplicate, stale, or mis-targeted bounty submissions. Must align agent guide, bounty rules, or PR template with claim-window expectations, live bounty capacity checks, active attempts, evidence requirements, and public-safety wording. Must include docs smoke evidence. Duplicate wording tweaks, broad rewrites, marketing copy, or unrelated docs churn do not qualify.
Focused MCP submit_work_proof structured guidance improvements without breaking text callers. Must include focused MCP tests, keep output grounded in actual application fields, and avoid duplicate, docs-only, broad rewrite, or unrelated discovery changes.
Focused code-health improvement that reduces real complexity, with priority on extracting coherent app/main.py subsystems or consolidating duplicated validation without public behavior changes. Must include targeted regression tests, avoid broad rewrites or formatting churn, and preserve API, MCP, ledger, wallet, webhook, admin, and page behavior unless fixing a documented bug.
Review open MergeWork PRs with specific evidence. Accepted review claims must be by someone other than the PR author, link a GitHub PR review or concise PR comment, include concrete evidence, and be distinct from already paid review claims. Duplicate, vague, self-review, rubber-stamp, misleading, or superseded duplicate reviews do not qualify.
Reusable automation for maintainers to summarize open PR queue health by bounty issue, merge state, stale or closed bounty reference, needs-info, missing bounty reference, and likely duplicate scope. Must be non-destructive, fixture-testable, documented, and not a one-time triage report.
Advisory active-attempt reservation path for bounty work that reduces duplicate agent effort without creating payments, claiming acceptance, mutating ledger balances, or blocking maintainer decisions. Must include deterministic expiration behavior, visibility, and tests for duplicate active attempts, expiration, release or status update behavior, closed or exhausted bounty handling, and preserved payout behavior.
Focused code-health improvement that reduces real complexity, with priority on extracting coherent app/main.py subsystems or consolidating duplicated validation without public behavior changes. Must include targeted regression tests, avoid broad rewrites or formatting churn, and preserve API, MCP, ledger, wallet, webhook, admin, and page behavior unless fixing a documented bug.
Reusable pre-submission quality gate for agents or contributors that checks bounty status, required PR evidence, stale or exhausted bounty references, and duplicate or similar open PR risk without claiming acceptance or payment. Must include focused tests and preserve existing payout behavior.
Focused read-only browser admin UI for existing webhook event outcomes using admin cookie auth, status and limit filtering, safe displayed fields, and tests. Token-only API use, mutation controls, payload bodies, secrets, unrelated payout changes, or broad dashboards do not qualify.
Focused MCP submit_work_proof structured guidance improvements with format=json, backward-compatible text behavior, generic no-selector structured output, validation, and tests. Duplicate, docs-only, unrelated discovery, broad rewrites, or invented payment/claim state do not qualify.
Accepted review claims must link a GitHub PR review or concise PR comment, be by someone other than the PR author, include concrete evidence, be distinct from already paid claims, and avoid duplicate, vague, self-review, rubber-stamp, misleading, or superseded duplicate reviews.
Focused public-facing enhancements that help contributors find bounties, inspect accepted work, or understand proof/account activity, with tests. Duplicate, marketing-only, docs-only, broad redesign, or unrelated changes do not qualify.
Distinct useful PR reviews or PR comments by someone other than the PR author. Accepted reviews include concrete evidence, specific inspection notes, commands, CI checks, or actionable findings. Duplicate, vague, self-review, rubber-stamp, misleading, or superseded duplicate reviews do not qualify.
Useful distinct non-security bug report, small focused fix, or regression test with concrete repro evidence. Private security details must stay private until approved. Duplicate, vague, misleading, no-value, broad rewrite, unrelated, or private-detail submissions do not qualify.
Focused public-facing enhancements that help contributors find bounties, inspect accepted work, or understand proof/account activity, with tests. Duplicate, marketing-only, docs-only, broad redesign, or unrelated changes do not qualify.
Focused MCP/API behavior or validation improvements for agent bounty discovery, state inspection, proof lookup, or work-proof guidance, with tests. Docs-only API examples from #229, invented fields, broad rewrites, or marketing copy do not qualify.
Focused webhook or admin payout observability improvements with tests for idempotency, duplicate delivery, exhausted/already-paid cases, invalid links, payout notes, or proof linkage. Generic typo fixes, unrelated UI changes, or broad rewrites do not qualify.
Focused fixes or tests that harden GitHub wallet linking and GitHub-balance claim behavior. Must preserve nonce, signature, replay, and supply rules. Generic wallet validation and docs-only changes do not qualify unless tied to tested claim/link behavior.
Focused maintainer reconciliation or duplicate/payment audit improvements with tests. Must cover accepted-unpaid, duplicate source, exhausted overflow, stale labels, proof/source mismatch, or claim-source gaps. Generic docs-only, broad refactors, or duplicate reports do not qualify.
Focused public documentation PRs that make API or MCP examples match actual MergeWork response shapes, with evidence and docs/tests. Duplicate, invented, stale, style-only, or unrelated changes do not qualify.
Useful distinct non-security bug report, small focused fix, or regression test with concrete repro evidence. Private security details must stay private until approved.
Distinct useful PR reviews or PR comments by someone other than the PR author. Accepted reviews include concrete evidence, specific inspection notes, or actionable findings. Duplicate, vague, self-review, rubber-stamp, misleading, or superseded duplicate reviews do not qualify.
Focused public-facing enhancements that help contributors find bounties, inspect accepted work, or understand proof/account activity, with tests. Duplicate, marketing-only, docs-only, broad redesign, or unrelated changes do not qualify.
Focused PRs that add or tighten one concrete deploy configuration validation with regression tests and preserved deploy gates. Duplicate, broad, coverage-reducing, or unrelated changes do not qualify.
Focused public documentation PRs that make API or MCP examples match actual MergeWork response shapes, with evidence and docs/tests. Duplicate, invented, stale, style-only, or unrelated changes do not qualify.
Useful distinct public smoke checks tied to deploy ae2e0a6 or latest payment and bounty changes. Duplicate, vague, non-reproducible, private, or unrelated checks do not qualify.
Distinct useful PR reviews or PR comments by someone other than the PR author. Accepted reviews include concrete evidence, specific inspection notes, or actionable findings. Duplicate, vague, self-review, rubber-stamp, misleading, or superseded duplicate reviews do not qualify.
Build a read-only public activity dashboard and API for proof-backed bounty payments. Include only bounty_payment entries with bounty_payment proofs; exclude github_claim, wallet_transfer, bounty_reserve, bounty_release, and genesis entries; do not merge identities or fetch GitHub live data; update tests and docs.
Useful distinct non-security bug report, small focused fix, or regression test with concrete repro evidence. Private security details must stay private until approved.
Focused public docs, templates, examples, or onboarding improvement with clear confusion or missing-step evidence and passing docs smoke/relevant checks.
Useful distinct public smoke check or concise verified report against live pages, API, MCP, wallet flow, or docs.
Useful PR review by a non-author with specific evidence. Duplicate, vague, wrong, or self reviews do not qualify.
Focused CI, deploy-readiness, or test reliability improvement with preserved safety checks.
Focused public UX, API, docs, or smoke-check fix with tests and passing checks.
Review a PR you did not author and comment with a useful review link. Review must include an actionable finding or clear no-blockers evidence with files, behavior, commands, or CI checked. Duplicate, vague, or author/self reviews do not qualify.
Review a PR you did not author and comment with a useful review link. Review must include an actionable finding or clear no-blockers evidence with files, behavior, commands, or CI checked. Rollover claims from issues #47 and #92 may qualify if unpaid and not duplicate, vague, or author/self reviews.
Review a PR you did not author and comment with a useful review link. Review must include an actionable finding or clear no-blockers evidence with files, behavior, commands, or CI checked. Rollover claims from issue #47 after its final accepted slot may qualify if unpaid and not duplicate, vague, or author/self reviews.
Accepted PR preserves or improves existing AGENTS, pytest, ruff, mypy, deploy readiness, docs smoke, and Docker gates; includes clear evidence; does not reduce coverage.
Useful distinct non-security bug report, small focused fix, or regression test with concrete repro evidence. Private security details must stay private until approved.
Review a PR you did not author and comment with a useful review link. Review must include an actionable finding or clear no-blockers evidence with files, behavior, commands, or CI checked.
Focused PR fixes one small public UX, docs, API, responsive, or smoke-check issue with targeted tests and without reducing CI/deploy coverage.
Useful public smoke checks or concise reports against live pages, API, MCP, wallet flow, or docs. One award per distinct useful verified check.
Focused PR adds a get_proof MCP tool with found and missing tests plus short agent-facing docs. Proof payloads and ledger behavior remain unchanged.
Focused PR adds API regression tests for paid and closed multi-award bounties in list/detail responses and active bounty counts.
Focused PR adds a tested reconciliation check that reports accepted submissions without matching ledger payments and already-paid cases without modifying production data.
Focused PR adds GitHub Actions CI for pytest, ruff format check, ruff check, and mypy on PRs and main pushes. No deployment, secrets, or release steps.
Focused PR improves ledger or proof explorer scanning for bounty payment and release entries without changing ledger data, proof payloads, balances, or hash behavior. Existing checks pass.
Focused PR improves public API or MCP examples with useful live host references, short prose, and passing checks. No secrets, exploit details, deployment changes, or price claims.
Focused PR improves wallet registration, GitHub linking, claim, or transfer clarity without changing wallet security behavior. Existing checks pass.
Focused PR adds tests for at least one meaningful multi-award edge case not already covered. Existing checks pass and no private keys, secrets, exploit details, deployment changes, or price claims are included.
Review a PR you did not author and comment with a useful review link. Review must include an actionable finding or clear no-blockers evidence with files, behavior, or commands checked.
Star the repository, create/register an MRWK wallet, sign in with GitHub, link the wallet, and comment with a mrwk1 address plus one concise note. No private keys, secrets, exploit details, deployment changes, or price claims.
- Open a focused PR that improves ledger entry or proof page readability, especially hashes, account links, proof hash, and payout metadata. - Do not change ledger data, proof payloads, or ledger behavior. - Existing project checks pass: `pytest`, `ruff format --check .`, `ruff check .`, and `mypy app`. - Do not include private keys, secrets, exploit details, deployment changes, or price claims.
- Open a focused PR that adds tests proving malformed wallet registration, transfer, link, or claim requests return controlled 4xx responses rather than 500s. - Minimal code fixes are acceptable if the new tests expose a bug. - Existing project checks pass: `pytest`, `ruff format --check .`, `ruff check .`, and `mypy app`. - Do not include private keys, secrets, exploit details, deployment changes, or price claims.
- Open a focused PR that improves the bounty detail page presentation for status, reward, issue link, and acceptance text. - Keep copy short, direct, and human. - Avoid marketing copy, price claims, and layout-heavy redesigns. - Existing project checks pass: `pytest`, `ruff format --check .`, `ruff check .`, and `mypy app`.
- Open a focused PR that adds tests for at least two wallet transfer failure cases. - Good cases include invalid nonce, wrong signer, insufficient balance, invalid recipient, excessive memo, or malformed amount. - Existing project checks pass: `pytest`, `ruff format --check .`, `ruff check .`, and `mypy app`. - Do not include private keys, secrets, exploit details, deployment changes, or price claims.
- Star the `ramimbo/mergework` repository. - Create and register an MRWK wallet public key. - Sign in with GitHub and link the wallet from `/me`. - Comment on this issue with the `mrwk1...` wallet address and one concise note about any confusing step. - Do not post private keys, seed material, secrets, exploit details, deployment changes, or price claims.
Deployment smoke confirms bounty reserve, accepted webhook, payment, and proof lookup.