Inventory
Identify formats and code-bearing files. Hash the exact original binary without executing it.
inspect · inventory · pinA practical, standard-library-powered way to inventory a game, pin the original build, map functions to Rust and compare deterministic state traces. A passing test is a bounded result, not a whole-game compatibility claim.
Give the agent a stable source of evidence and a next actionable gap.
Identify formats and code-bearing files. Hash the exact original binary without executing it.
inspect · inventory · pinMap original entry points to Rust functions and distinguish implemented, wired and trace-compared results.
ledger add · set · auditRun a deliberately selected original-side probe and Rust probe with identical scripted inputs.
oracle_runner.py --compareReject empty traces and stale evidence. Compare actual events with exact numbers by default.
parity · report · auditThe CLI requires no REA account or GPU. Ghidra, REA and Unicorn remain optional external tools used only when they improve the specific investigation.
python3 scripts/workbench.py doctor python3 scripts/workbench.py pin \ --original /games/Game.exe \ --output reference.lock.json python3 scripts/oracle_runner.py \ scenario.json \ --output ./capture \ --allow-execute-original --compare python3 scripts/workbench.py ledger \ --db port-ledger.sqlite audit
These commands show the interface; they do not claim a game has been tested here.
Separate environments. Both processes get distinct basic scratch folders. This is not a security sandbox. Run untrusted programs only in a VM.
Scope the claim. Lock file hashes and passing traces prove only a specific recorded comparison. They do not establish that a trace was independently captured or a game is 1:1 complete.
Preserve the original. Game assets, binaries and decompiler dumps stay on the user's local machine, outside Git history.