Hold the Flood — Tasks and Evidence

Maintainer: PM. Updated: 2026-09-06 UTC. Workflow: Ready → In progress → In review → Accepted, with Blocked and Cancelled available. Only PM marks Accepted. Proposed staffing does not mean an agent is running.

Current work

ID Outcome Owner State Dependency Acceptance evidence / reviewer
PM001 Establish hub, brief, tasks, and copy-ready prompts PM Accepted None Hub, Brief, Team and Tasks published; page contents and navigation checked by PM
OWN001 Establish game intent and known constraints Project owner, facilitated by PM Ready None Owner answers recorded with unresolved points clearly marked / PM
DES001 Propose the smallest test of the intended game Game designer (not created) Blocked Agent creation and core owner brief One-page interaction specification; choices, feedback, success/failure, three design risks, explicit cuts / PM plus QA critique
ENG001 Establish existing implementation and run constraints Prototype engineer (not created) Blocked Agent creation; relevant project access or confirmation none exists Verifiable inventory; environment limits; build path; smallest technical risk experiment / PM
QA001 Challenge the prototype proposal and define checks Playtest and QA reviewer (not created) Blocked Agent creation; brief; DES001 proposal Specific falsification questions, observable checks and severity rules / PM
ENG002 Produce a playable core-loop experiment Prototype engineer (not created) Blocked PM-scoped design, workable environment, reviewed plan Runnable artifact and revision; controls; meaningful choice; visible consequence; conclusion/retry; observed run evidence / QA then PM
QA002 Independently review the exact prototype build Playtest and QA reviewer (not created) Blocked ENG002 artifact Run record; reproducible issues; untested areas; gate recommendation / PM
OWN002 Play and judge the direction Project owner, facilitated by PM Blocked Runnable reviewed prototype Owner feedback on intended experience; iterate/pivot/advance decision / PM records

DES001's first brief may be short. ENG001 can proceed alongside design once access is available. Do not wait for a large design document.

Task detail template

  • ID and title:
  • One-sentence player or project outcome:
  • Implementer / independent reviewer:
  • State / last meaningful update (UTC):
  • Dependencies and required access:
  • Deliverable location and revision:
  • In scope / explicit cuts:
  • Acceptance criteria (observable):
  • Verification performed and result:
  • Known risks, blocker, next action:
  • Review verdict, PM acceptance, and supporting links:

First playable gate

Apply to the approved game concept and platform; adjust only with an explicit task decision.

  • A reviewer can obtain and start the specified revision using the supplied instructions.
  • The player can discover the controls and goal.
  • The intended core action presents at least one meaningful choice with readable consequences.
  • The short test has a clear end or outcome and can be repeated.
  • The exact build has been independently exercised. Critical crashes, softlocks, and loss of required progress block acceptance.
  • The owner has a playtest question to answer. No invented claim of enjoyment or user validation.

Review and rework

A rejection names the artifact revision, failed criterion, observation or reproduction steps, severity, and smallest useful correction. Engineer addresses the issue; reviewer checks the changed behavior. Avoid circular debate and unlimited polishing. PM decides whether another attempt is justified or the scope should change.

Current verified output: wiki inspection and initial process setup only. No game agent has been created, and no playable build has been reviewed.

0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9