# Vgrichina Re Skill Re Skill > Live context - Skill: `tomevault-io/vgrichina-re-skill-re-skill` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/vgrichina-re-skill-re-skill` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/vgrichina-re-skill-re-skill/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Coding & Dev Tools - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/vgrichina-re-skill-re-skill --- ## Live context - Open tasks: !`grep -m 10 '\[ \]' REVERSE.md 2>/dev/null || true` - Dead ends: !`head -50 dead_ends.md 2>/dev/null || true` - Sessions: !`ls re_loop_sessions/ 2>/dev/null | tail -5 || true` ## Dispatch **$ARGUMENTS given -> bootstrap new project** - Read [phases.md](phases.md), [reverse_template.md](reverse_template.md), [dead_ends_template.md](dead_ends_template.md), [re_loop_template.sh](re_loop_template.sh) - Scaffold: REVERSE.md, labels.csv, dead_ends.md, tools/ dir, re_loop.sh (fill config block) - Identify binary header -> fill Ph1 in REVERSE.md - Add first tasks to ## Next Tasks -> ### RE Investigation **no $ARGUMENTS -> continue from REVERSE.md** - Read CLAUDE.md for platform conventions, tool prefix, binary path - Pick top unchecked task from ## Next Tasks - Skim re_loop_sessions/*.txt (last 3) to avoid repeating work - Read dead_ends.md to avoid known dead-end approaches ## Task categories **RE investigation** (tools/dis.py, xref.py, search_bytes.py, decode_tables.py) 1. Run 2-3 tool calls 2. Write findings to REVERSE.md immediately — do NOT batch all investigation before writing 3. Add new addresses to labels.csv (addr,name,comment) 4. Repeat: more tool calls -> write, until task fully understood 5. Mark [x], add new [ ] discoveries **Web port fix** (web/ files) 1. Read the relevant web/ file first 2. Apply fix — values are already documented in REVERSE.md 3. Update REVERSE.md to note what changed 4. If new asset type decoded -> add section to web/catalog.html **Documentation** (docs/ directory) 1. docs/architecture_exe.md — binary layout, segment/bank boundaries, call graph with entry points + file offsets, module map, per-turn execution flow, data flow, shared data regions 2. docs/architecture_web.md — JS module graph with exports + dependencies, EXE->web mapping table, game loop flow, state machine with transitions Both files must exist before the project is considered complete. ## Knowledge base **labels.csv** accumulates across sessions. All disasm tools load it automatically. - Flat (DOS/unpacked): `offset,name,comment` — offset with 0x prefix e.g. `0x25DE9,main,entry point` - Banked ROM (NES/GB): `bank,addr,name,comment` — addr hex no prefix e.g. `0,C070,Reset,` - DS-relative (DOS): `DS:offset_hex,name,comment` e.g. `DS:0xCEAC,dt_physics,timestep` The comment column is optional but encouraged for non-obvious addresses. ## Stuck rule After 10 tool calls with no progress on a single task: 1. Append entry to dead_ends.md: tried / why-failed / better-approach / session number 2. Split stuck task into 2-3 smaller sub-tasks in REVERSE.md Next Tasks 3. Mark original [x] "Split into sub-tasks" — move on If the same task is attempted across 3+ sessions without progress, escalate: the task decomposition is wrong, rethink the approach entirely. ## Common gotchas (from past projects) - **MZ relocations break byte-pattern searches**: raw `search_bytes.py` misses far calls/jumps because segment values are patched at load time. Use `xref.py` which understands relocation tables. - **Decompress before disassembly**: compressed binaries are opaque. Always identify and unpack first (Ph2). All subsequent work uses the unpacked image. - **Two-level pointer tables**: flat decode shows pointer values, not the data they point to. Use `decode_tables.py --follow` to chase pointers to sub-tables. - **Tile index ≠ VRAM address**: many platforms use attribute/remapping tables between tile indices and actual video memory. Validate tile rendering against emulator, don't assume direct mapping. - **Don't search for composed text**: menu text, scores, and dialog are often built at runtime from format strings + data tables. Trace from code control flow, not string patterns. - **Bank-aware xref**: on banked ROMs (NES/GB), cross-bank byte matches are false positives. Always pass the bank filter to `xref.py`. - **WRAM tilemap stride varies**: Game Boy WRAM buffers may use 20-col stride, not the standard 32-col VRAM stride. Verify actual layout. - **Struct stride off-by-one**: a wrong stride corrupts every subsequent struct in the table. Cross-check decoded fields against emulator memory dump. ## Tool rules - Read not cat · Grep not grep · Glob not ls/find - Write a .py file first, then run it — never inline `python3 -c` - No third-party disassemblers (no radare2, Ghidra, ndisasm, mgbdis) - All tool scripts live in tools/ — invoke as `python3 tools/dis.py ...` ## Tool CLI reference All tools auto-load labels.csv for annotation. Prefix is `python3 tools/`. Core tools (signatures adapt per platform): ``` dis.py [lines] # flat binary: file offset (hex) dis.py [nbytes] # banked ROM: bank + addr dis.py [nbytes] --hunk N # Amiga hunk: target specific hunk xref.py # find all refs (calls, jumps, loads, stores) xref.py [bank] # banked: optional bank filter xref.py --code # scan known code segments only (faster) xref.py -r START END # restrict scan to file range xref.py -c N # show N bytes of context search_bytes.py [--context N] [--disasm [N]] # hex can be spaced or not decode_tables.py # fmt: u8/s8/u16/s16/u32/ptr16/farptr/q8/b8/nullstr decode_tables.py struct:: # struct with field layout decode_tables.py --follow N FMT # two-level pointer tables extract_tiles.py # decode tiles/sprites to gfx/ PNGs render_screen.py # composite screen to gfx/screen_NNN.png ``` Platform extras (built as needed): ``` decompress.py # reverse-engineered decompressor (Ph2) seg_offset.py # convert between SEG:OFF and file/DS offsets (DOS) ds_lookup.py [-n N] [-s|-w|-d|-f32|-f64] # DS-relative memory viewer strings_dump.py [-g "pattern"] [-r START END] # string scanner with grep/range struct_dump.py # pre-configured struct dumper find_callers.py # find all callers (near+far), complements xref.py palette_dump.py [--accent] [--scan] # palette extractor a4resolve.py [--dump | -- ] # Amiga A4-relative library resolver hunk_scan.py / hunk_loader.py # Amiga hunk executable parser adf_extract.py / adf_debug.py # Amiga ADF disk image tools check_bank_refs.py # verify far references point to valid banks (NES/GB) extract_rom_banks.py # split MBC1/3 ROM banks (GB) render_sprites.py # render sprite sheets with palette/scale options render_level.py # full level/map renderer (Ph6) render_compare.py IMG1 IMG2 [--diff] # side-by-side pixel comparison compare_frames.py # pixel-diff emulator vs web port (Ph6) ``` Scriptable emulator (Ph5.5, `python3 -m tools.emu` or `python3 tools/emu`): ``` --dump-regs # load binary, print initial state --boot-test # run until first OS call or N instructions --run-func LABEL # jump to labeled function with pre-set state --break ADDR # breakpoint, dump registers + stack --trace # print each instruction --keys STEP:SCANCODE:ASCII # inject key input at step N --dump-screen FILE.png # screen capture --compare # export CSV for diffing against web port ``` ## Scaffolded project structure ``` REVERSE.md # data-range map, findings, task list, verification checklist labels.csv # addr,name,comment — grows across sessions dead_ends.md # stuck log — avoids repeating failed approaches re_loop.sh # autonomous session driver re_loop_sessions/ # session logs tools/ instruction_set.py # CPU opcode database (built per platform) dis.py # targeted disassembler (auto-loads labels.csv) search_bytes.py # byte pattern search xref.py # cross-reference finder decode_tables.py # struct/table decoder extract_tiles.py # graphics decoder render_screen.py # screen compositor emu/ # scriptable emulator (Ph5.5, optional) gfx/ # extracted graphics web/ index.html # game reimplementation catalog.html # asset browser for visual validation docs/ architecture_exe.md # binary layout, call graph, data flow architecture_web.md # JS modules, EXE-to-web mapping ``` ## Platform conventions Address notation and labels.csv format vary by platform. Set in CLAUDE.md during bootstrap. - **DOS flat/MZ**: file offsets `0xXXXXX`, DS-relative `DS:0xXXXX`, segment:offset `SEG:OFF`. Tools: seg_offset.py, ds_lookup.py. FPU emulation (INT 34h-3Dh) if Borland. - **NES/GB banked ROM**: `bank,addr` (addr hex no prefix). WRAM/HRAM labels use `bank='*'`. Tools: check_bank_refs.py, extract_rom_banks.py. - **Amiga hunk**: hunk-relative offsets, A4-relative library calls. Tools: hunk_scan.py, a4resolve.py, adf_extract.py. ## re_loop.sh (scaffolded for user, not agent-invoked) Scaffolded into the project from [re_loop_template.sh](re_loop_template.sh) during bootstrap. The user runs it from the terminal to drive autonomous sessions — the agent never invokes it directly. The ALLOWED_TOOLS in re_loop.sh should include wildcard patterns for on-the-fly scripts: `Bash(python3 tools/analyze_*)`, `Bash(python3 tools/audit_*)`, `Bash(python3 tools/check_*)`, `Bash(python3 tools/gen_*)`, `Bash(python3 tools/dump_*)`. ## Verification checkpoints Mark these in REVERSE.md as validation happens: - [ ] Ph3: 3+ functions traced and cross-checked against emulator trace - [ ] Ph4: 5+ sprites/tiles extracted and visually compared to emulator - [ ] Ph5: key data struct confirmed in emulator memory dump, all fields match - [ ] Ph6: full game session played, no major logic gaps found - [ ] Ph7: web port pixel-compared against emulator screenshots ## End of session - Mark completed task [x] - Add newly discovered unknowns as [ ] in the appropriate ### sub-section - Last line: `SESSION_SUMMARY: ` See [phases.md](phases.md) when bootstrapping a new project. --- > Source: [vgrichina/re-skill](https://github.com/vgrichina/re-skill) — distributed by [TomeVault](https://tomevault.io).