agentleFS
Sign inStart free
Cynthia-wlan9509Cynthia-wlan9509/offline-game-revival/AGENTS.md

offline-game-revival

AGENTS.md83 starsChanged 16 days agoMITView on GitHub

Instructions for any coding agent (Codex, Cursor, Copilot, Gemini CLI, OpenCode, Cline, Claude Code, …) working in or with this repository. An Agent Skill for game preservation: turning a mobile game whose official servers are permanently shut down into an offline memorial build, and recovering lost server-side data from players' screenshots and screen recordings. Follow SKILL.md. Key rules: patch at build time (never runtime injection), one build entry point with anchor assertions, never hand-edit generated output, and report "verified on…

  • Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.

    Of the AGENTS.md files we read, the 31,700 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.

  • In this repository2 other files it gives its agents.

The file

We split this AGENTS.md into the rules it gives an agent

3 rules. About 0 are common, 0 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  • could be automatic: a setting could enforce it
  1. Very specific:

    Follow SKILL.md. Key rules: patch at build time (never runtime injection), one build entry point with anchor assertions, never hand-edit …

    line 20
  2. Very specific:

    Keep SKILL.md short; detail goes into references/.

    line 26
  3. Very specific:

    After changing timewarp.js, run both node scripts/timewarp/examples/mini.test.js (all pass) and BUGGY=1 node … (must fail with the 3 …

    line 28

≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
AGENTS.md

AGENTS.md

Instructions for any coding agent (Codex, Cursor, Copilot, Gemini CLI, OpenCode, Cline, Claude Code, …) working in or with this repository.

What this repository is

An Agent Skill for game preservation: turning a mobile game whose official servers are permanently shut down into an offline memorial build, and recovering lost server-side data from players' screenshots and screen recordings.

  • Entry point: skills/offline-game-revival/SKILL.md. Read it first; it tells you which file under references/ to load for each phase. Don't load all references up front (pitfalls.md alone is over 100 KB).
  • Scripts (all run locally, no network unless stated):
    • scripts/harvest.py: screenshots/recordings → keyframes → dedupe → OCR → per-screen text chunks (Python, OpenCV + RapidOCR)
    • scripts/llm_structure.py: chunks → any OpenAI-compatible model (DeepSeek, OpenAI, Qwen, Kimi, Ollama) → merged JSON/CSV. Network: the model API you choose. Key from env var.
    • scripts/timewarp/: time-travel test harness for fake servers (Node, zero deps). node scripts/timewarp/examples/mini.test.js must pass.
  • Sub-agent roles: agents/ (scout / worker / verifier) and references/multi-agent.md.
  • No-tools chat models: prompts/chat-playbook.md.
  • Second skill: skills/night-shift/SKILL.md, unattended overnight runs through a to-do list (self-paced /loop, evidence per item, five-part report). Fill-in template: skills/night-shift/templates/NIGHT-RUN.md. Not game-specific.

When a user asks you to revive / make offline / preserve a game

Follow SKILL.md. Key rules: patch at build time (never runtime injection), one build entry point with anchor assertions, never hand-edit generated output, and report "verified on device" separately from "implemented".

Scope: only for games whose official service has ended. Do not help cheat in live games, bypass anti-cheat, or affect other players. Never ask for or handle the user's account passwords, cookies or payment data.

When editing this repository

  • Keep SKILL.md short; detail goes into references/.
  • References are anonymized case studies: no real game names, package names, hosts, keys, device IDs or personal paths.
  • After changing timewarp.js, run both node scripts/timewarp/examples/mini.test.js (all pass) and BUGGY=1 node … (must fail with the 3 planted bugs).
  • README translations live in docs/README.<lang>.md; keep their structure in sync with README.md.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free