Skip to content

Roadmap

clockwork is under active development against a live game client. This page is a snapshot of what's shipped, what's in progress, and what's next — not a promise of dates. If a feature you're relying on shows up here as unfinished, that's the state of it; if a page elsewhere on this site describes something that contradicts this one, this page is more likely to be stale — check the project's issue tracker for the current word.

For permanent design choices (not gaps to be filled in) see Differences from MacroQuest instead.

Marker Meaning
Shipped Live and usable today.
In progress Partially working, actively being built out.
Planned Scoped or designed, not started.
Blocked Waiting on something else landing first.

Combat automation (jidou)

The biggest single item on this roadmap: jidou has no combat engine yet. It's a complete chassis — script host, targeting scaffolding, per-class config format, a working pull module — but the piece that actually runs a rotation (cast primitives, a targeting/combat state machine, a rotation executor) isn't wired in. Until it lands, jidou automates positioning and support tasks but does not fight.

What Status
Class-independent combat engine (casting, targeting, rotation executor) Planned
Pulling module — full parity (unified camp logic, multiple pull modes, peer vitals watch) In progress
Buff-starvation fix across several shipped class configs (an over-broad "self-maintained buffs" list can starve rotation rows below it) In progress
Autoloot — pipeline works end to end, blocked on knowing free bag space before looting Blocked
jidou UI parity with RGMercs (tabbed window, HUD mode, target window, theming, 2D map) — today it's a flat text window Planned
Item/clicky use inside rotations Planned
A handful of specific per-class rotation gaps (e.g. a Fury precast chain, Swashbuckler flanking detection) Planned
What Status
Mesh-free waypoint navigation (offline-baked .jbm zone data) — early phases validated In progress
Navmesh/path overlay + click-to-navigate — mesh, path, and point-of-interest overlays are live; drawing them over the real (non-top-down) camera view needs camera-matrix access clockwork doesn't have yet In progress
Swimming/flying movement (needs different handling than ground movement) Planned
Full raid-roster detection — the client's raid-subgroup layout in memory hasn't been located yet, which caps how complete Raid and raid-scoped multibox broadcasts (/dgra//dgre) can be Blocked

Multibox

What Status
Background hotbar/key broadcast across peers — the full input-relay pipeline is built and reverse-engineered, but hasn't been run against a live client yet In progress
A general cross-client messaging bus — /dga//dge are wired directly for now and will become this bus's first built-in user once it lands In progress
Raid-scoped broadcast reliability Blocked — see raid-roster detection above

UI automation

What Status
<List>-control row reading (unlocks quest-journal picks, group/raid panel rows, and more) — mostly scoped, close to landing Planned
Recursive UI-tree inspector with a live property grid Planned
2FA / security-challenge login flows — normal auto-login works, these don't yet Planned
Console color-token support (\aw/\ag-style chat coloring) Planned

Lua API

What Status
A key-binding system — react to a real physical keypress, not just polling inside an ImGui frame Planned
ImGui.Image / ImageButton — currently no-op placeholders that reserve layout space and draw nothing, since binding a raw texture id the wrong way is a guaranteed client crash Planned
ImPlot bindings Planned

Core / engine

What Status
Crash minidump pipeline — a crash logger exists today; a full Crashpad-style minidump doesn't yet In progress
Central game-thread work queue — the foundation later phases are built on In progress

Known issues

  • A Lua crash tied to stale script state, roughly half a second after character death. Open, not yet root-caused.
  • A rare client crash at character select, triggered by a connection-teardown race condition.

Non-goals

Explicitly out of scope, not just "not done yet":

  • No C-ABI plugin interface. Lua is the only extension surface — see Differences from MacroQuest.
  • No .mac macro language.
  • No committed cross-PC/LAN multibox target. Same-box multibox is what's built and tested; the design doesn't rule out LAN, but it isn't a promise.
  • No EverQuest II XTarget equivalent — the game itself has no such window.
  • No anti-cheat evasion or hardening shipped. EverQuest II has no anti-cheat today; this would only be revisited if that changes.

See also