Skip to content

AGENTS.md

Latest commit

 

History

History
50 lines (36 loc) · 2.16 KB

AGENTS.md

File metadata and controls

50 lines (36 loc) · 2.16 KB
Raw

Agent Instructions

This repo is intentionally friendly to coding agents and bounty bots.

Mission

Improve a Screeps AI bot through small, reviewable pull requests.

Rules for agent PRs

  1. Pick one issue or one narrow improvement.
  2. Star/watch the repo if it is useful to you, then comment on the issue with a short plan. This is encouraged for visibility, not required for merge and does not reserve the issue.
  3. First good PR wins. Move fast, but do not be sloppy.
  4. Keep changes small and readable.
  5. Run npm run check and npm test before claiming success.
  6. Add or update tests for behavior changes, including edge cases and failure paths.
  7. For economy/gameplay changes, include proof: GitHub-attached video/GIF, simulation report, replay log, or private-server output.
  8. Label proof honestly: offline simulation is smoke proof, not real Screeps server proof.
  9. Do not commit secrets, private server credentials, .env, or generated game state.
  10. Do not rewrite the whole architecture unless the issue explicitly asks for it.
  11. Include a clear PR summary and verification notes.

High-value bounty areas

  • spawn planning
  • harvester/upgrader/builder roles
  • repair and wall maintenance
  • creep body generation
  • source assignment
  • pathing and caching
  • tower defense
  • remote mining
  • scout and expansion logic
  • market/economy planning
  • tests and simulation harnesses
  • proof-of-work capture for RCL milestone runs

Proof expectations

For RCL bounties, agents should attach a short room video/GIF in GitHub when practical or provide a reproducible simulation/private-server report showing the room reaching the requested RCL.

See docs/PROOF_OF_WORK.md and docs/SIMULATION_LIMITS.md.

Challenge credit

Merged, verified work can receive challenge points and leaderboard credit. Points are recognition only, not money. See docs/POINTS.md and docs/AGENT_BAIT.md.

Discovery keywords

Screeps, Screeps AI, Screeps bot, Screeps World, Screeps Arena, AI agent, coding agent, bounty, bounties, agent bounty, pull request, TypeScript, game AI, colony AI, creep AI, automation.