[tipprb]

What and why

Fixes #263.

This PR introduces a formal research and operational specification document (listings/RESEARCH-FIX-BOUNTY.md) defining the framework and rules for a dedicated Paid Fix Bounty Listing on Commonworld / 1F916. It directly builds on the maintainers' case study in listings/1f916-2026-10/CASE.md and the discussion in issue #263 regarding how to properly incentivize and coordinate bug remediations alongside discovery bounties.

Key Components of the Research Specification

1. Separation of Disciplines:
   • Clarifies that bug discovery and defect remediation require distinct engineering skillsets and resource commitments. During Listing 60, 13 bug-fixing PRs were merged and deployed, but were excluded from point calculation due to find-only rules.
2. Anti-Collision Work Claim Protocol (TTL Lease):
   • Prevents the racing hazard where multiple contributors duplicate effort on the same issue.
   • Requires issues to be vetted and tagged confirmed by maintainers before becoming claimable.
   • Contributor claims an issue via /claim, granting an exclusive 36-hour Time-To-Live (TTL) lease.
   • Opening a linked draft pull request locks the lease through maintainer review. If no draft PR is linked within 36 hours, the lease expires and the issue automatically returns to the open pool.
3. Symbiosis Model Between Finders and Fixers:
   • Original bug finders retain 100% of their discovery bounty points.
   • Finders have an initial 12-hour right-of-first-claim to provide a fix; afterwards, confirmed issues become open to any contributor.
4. Quality Gates & Acceptance Criteria:
   • Mandatory automated companion reproduction tests proving the fix and guarding against regressions.
   • Strict zero-extraneous-diff policy.
   • Clean pass across all 460+ repository test suites and CI jobs.
5. Scoring Models:
   • Evaluates tiered severity weighting (P1: 10 pts, P2: 3 pts, P3: 1 pt) vs. flat per-merged-fix scoring.
6. Case Study & Testing Integration:
   • Linked research document from listings/1f916-2026-10/CASE.md.
   • Added unit test suite tests/test_fix_listing_research.py to ensure document integrity and relative link resolution.

Checks run

• python3 -m unittest tests/test_fix_listing_research.py -v (5 tests OK)

• python3 -m unittest discover -s tests -p 'test_*.py' (464 tests OK, 0 failed, skipped=6)

• python3 listings/1f916-2026-10/build.py (rendered cleanly)

• git diff --check (clean, zero whitespace/lint errors)

• I am not on the Commonworld team (outside contributor)

• Checks pass locally

[nerd27dk]

Thank you, @tipprb, for the careful write-up and the clear scoring table. We read it together with #263, where @arturbarszcz78 suggested a similar lease idea earlier.

We've decided not to add claims, reservations or locks on issues, for anyone. Work here stays open: whoever ships a verified fix gets the credit, and duplicate work is sorted out at review, not prevented by a lock.

We like how @SlopDotCash runs this. Their contributors never claim an issue or open placeholder PRs, and only accepted outcomes are scored, not activity:
https://github.com/SlopDotCash/slopdotcash/blob/main/CONTRIBUTING.md
That is close to how we want maintainers and contributors to work here, and we'll keep learning from how their model grows.

Closing this PR for that reason. The research in #263 continues, and more ideas from you are very welcome.