New Your team’s decisions, in one playbook every coding agent works from. Never answer your agent twice

3do Dev

48 tools. 30 can modify or destroy data without limits.

30 write tools that can modify data. Rate limits recommended.

Last updated:

30 can modify or destroy data
18 read-only
48 tools total

Community server · catalogue entry checked 13/10/2026

How to control 3do Dev ↓

What 3do Dev exposes to your agents

Read (18) Write / Execute (30) Destructive / Financial (0)
High Risk

The most dangerous 3do Dev tools

30 of 3do Dev's 48 tools can modify, destroy, or commit something on every call — and an agent calls them with no built-in limits.

How to control 3do Dev

PolicyLayer is an MCP gateway — it sits between your AI agents and 3do Dev, and nothing reaches the server without passing your rules. These are the rules we recommend:

Rate limit write operations
{
  "emu_state": {
    "limits": [
      {
        "counter": "emu_state_per_hour",
        "window": "hour",
        "max": 30,
        "scope": "grant"
      }
    ]
  }
}

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations
{
  "emu_crashes": {
    "limits": [
      {
        "counter": "emu_crashes_per_minute",
        "window": "minute",
        "max": 60,
        "scope": "grant"
      }
    ]
  }
}

Controls API costs and prevents retry loops from exhausting upstream rate limits.

  1. Create a free account and register 3do Dev — nothing to install.
  2. Add these rules — paste them, or build them visually. Tune the limits to your setup.
  3. Point your MCP client (Claude, Cursor, anything) at your gateway URL.
ENFORCE POLICY ON 3DO DEV →

Instant setup, no code required.

All 48 3do Dev tools

EXECUTE 28 tools
Execute build Compile projects/<project> (Norcroft ARM C) and pack a signed ISO + ELF/symbols. Execute emu_attach Attach to a running session (e.g. a window started with ./3do run demo, control Execute emu_boot emu_boot Execute emu_break Set (or remove) a halting breakpoint at a function/symbol ("update_plasma", "main+0x20") Execute emu_cmd Any session control command, e.g. emu_cmd("help"), emu_cmd("find_mem", {"text": "DEMO"}), Execute emu_continue Resume a halted CPU and run until the next breakpoint/watchpoint or max_frames Execute emu_halt Stop the CPU now (before its next instruction); shows where. Execute emu_hold Hold buttons across steps/free-run until changed; "" releases. Execute emu_input Non-pad input. device: joypad|flightstick|mouse|lightgun|arcade_lightgun|orbatak_trackball|none Execute emu_look Optionally press BUTTONS, run FRAMES, then return new debug output + a screenshot. Execute emu_pause Stop free-running (frames then only advance on step/press). Execute emu_press Press+release pad buttons: "A", "UP+B", "P". Buttons: A B C P(play) X(stop) L R UP DOWN LEFT RIGHT Execute emu_reboot Power-cycle the console, optionally inserting another disc (project name or ISO). Execute emu_record_gif Run FRAMES frames capturing every Nth into an animated GIF (path relative to repo root). Execute emu_regs CPU registers with symbolized pc/lr and mode. set_name/set_value change one (e.g. r0, 0x10). Execute emu_reset Soft reset (console reset button). Execute emu_restart Kill and respawn the session process (same disc, same window setting). Use if it hangs. Execute emu_run Free-run continuously (speed 1.0 = real time, 0 = max). Also resumes a halted CPU. Execute emu_run_until Step until TEXT appears in debug output emitted after the last emu_log call. Execute emu_sequence emu_sequence Execute emu_step Advance N frames (60 = 1 s of 3DO time). Stops early on a breakpoint/watchpoint. Execute emu_stepi Execute COUNT machine instructions then halt; returns the new pc and symbol. Execute emu_stop End the current session (stops it if this server started it; detaches otherwise). Execute emu_trace Tracepoints (log args without stopping): action add|remove|clear|hits; plus action Execute emu_watch Data watchpoint on a global/symbol or address (kind: write|read|access); halts after Execute emu_write_mem Write bytes (hex) or one 32-bit value (u32="0x1234") at an address or symbol. Execute gdb_run Run gdb commands against the current session with full source-level symbols Execute run_tests Run a project's tests headless (./3do test <project>), or every project (./3do selftest).
READ 18 tools
Read emu_crashes Aborts / exceptions the 3DO OS reported, with registers and log context. Read emu_disasm Disassemble ARM code at pc, a symbol, or an address, annotated with symbols. Read emu_log Debug console (kernel messages, kprintf/printf/tdo_log). Default: new since last call; since=0 for all. Read emu_profile Statistical CPU profile over FRAMES frames: % of samples per function (OS/idle bucketed). Read emu_ps OS introspection from kernel memory. what: tasks (process list: state, priority, Read emu_read_mem Read memory at an address or symbol (e.g. a global). as_u32: decode big-endian words. Read emu_screenshot Current frame (320x240 x scale). save_as: also write a PNG (relative to repo root). Read emu_sessions List running emulator sessions on this machine (pid, iso, control/gdb ports, window). Read emu_status Game, frame, run mode, CPU halt state + location, held input, devices, ports. Read emu_symbols Search the program's symbols (regex) with runtime addresses. Read emu_syscalls System-call snoop. action: start | stop | fetch (counts by call name + latest calls Read list_projects List 3DO projects (projects/<name>) and their built ISOs. Read os_diff What changed since os_snapshot(name): net item count per type/owner (positive = possible Read os_inspect os_inspect Read os_item One OS item by number, with type-specific detail and its raw node bytes. Read os_memory_map DRAM/VRAM page map: which task group owns each page ('.' free, '?' kernel). Read os_overview Portfolio OS summary: kernel base, current task, live item counts by type, DRAM/VRAM used/free. Read os_snapshot Record OS state (items, tasks, free memory) to compare later with os_diff.

Related servers

Other MCP servers with similar tools — same risk classification, starter policies for each.

Questions about 3do Dev

How do I prevent bulk modifications through 3do Dev? +

The 3do Dev server has 2 write tools including emu_state, new_project. Set a rate limit in your policy -- for example, 10 calls per hour prevents an agent from making more than 10 modifications per hour. PolicyLayer enforces this at the gateway, before calls reach 3do Dev.

How many tools does the 3do Dev MCP server expose? +

48 tools across 4 categories: Destructive, Execute, Read, Write. 18 are read-only. 30 can modify, create, or delete data.

How do I enforce a policy on 3do Dev? +

Register the 3do Dev MCP server in PolicyLayer, apply the suggested rules above (adjust the limits to your use case), and point your AI client at the PolicyLayer proxy URL instead of the server directly. Your agents keep the same tools; PolicyLayer evaluates every call against policy before it executes. Nothing to install, live in minutes.

Enforce policy on every 3do Dev tool call.

Deterministic rules across all 48 3do Dev tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.

Instant setup, no code required.

48 3do Dev tools catalogued and risk-classified — across an index of 46,500+ MCP servers.

// WHERE THIS COMES FROM

These policies come from 3do Dev's registry record.

The record behind this page: verified identity, auth posture, risk grade, every tool classified, recommended policy — re-checked continuously.

Teams ship this data inside their own products. See what a licence covers →

// GET IN TOUCH

Have a question or want to learn more? Send us a message.

Message sent.

We'll get back to you soon.