| name | dev-loop |
| description | Use whenever you change any Lua module under Scripts/ee-dcs/ — the mandatory edit→build→check→inject→observe cycle for this EECH→DCS port, including the 0-warnings/no-findings gates, live re-injection, reading dcs.log, the verified-type-name DB guardrail, and the reserve-refund rule for spawner code. |
Dev loop: edit → build → check → inject → observe
The one true cycle for changing this port. Every change goes through it. Do not skip build or
check — they are gates, not suggestions.
0. Prerequisite: DCS Studio must be running
Build, static analysis (check), live eval/injection (dcs_eval), and live DB queries are not
shell commands — they are tools exposed by the DCS Studio MCP server at
http://127.0.0.1:25570/mcp (registered in .mcp.json). The DCS Studio desktop app must be
running for these tools to appear in a Claude session.
- Detect absence: if the build / check /
dcs_eval / DB-query MCP tools are not present in your
tool list, the app is not running (or the session predates its launch).
- What to tell the user: "Start the DCS Studio desktop app, then restart or reconnect this
Claude session so the MCP tools register." Do not try to fake it.
- Never run
dcs-studio from the shell to get help/output — that launches the full desktop app,
it does not give you the tools.
1. Edit
Edit the modules under Scripts/ee-dcs/. One system per module; each module header
names its EECH source file(s) — keep that header accurate (see the eech-fidelity skill).
2. Build (mandatory gate)
Run the DCS Studio build tool. It transpiles/bundles the 22 modules into
dist/ee-dcs.lua.
- Healthy output: 22 modules, 0 warnings.
- Any warning is a stop-the-line failure. Fix it before proceeding — do not inject a warning build.
3. Check (mandatory gate)
Run the DCS Studio check (static analysis) tool over all files.
- Healthy output: all files scanned, NO findings.
- Any finding is a stop-the-line failure. Fix it (or, if it is a documented deliberate proxy, that
belongs in
eech-fidelity — check does not normally flag those). Do not inject with findings open.
The 0-warnings / no-findings discipline is absolute: both gates green after every change, no
exceptions, before you inject.
4. Inject into the live mission
With a mission running (see the live-validation skill for what that mission must be), inject via
the dcs_eval MCP tool:
net.dostring_in('server', 'dofile("C:/Users/jonat/DCSStudio/my-test-mod/dist/ee-dcs.lua")')
Re-injection is safe mid-session. campaign_state.lua bumps a global _DMT_GEN (generation
counter) on every load; every scheduled closure captures my_gen = cs.GENERATION and self-cancels
(return nil) when cs.GENERATION ~= my_gen. So a hot re-inject supersedes the old timers instead
of double-running them. Caveat: this is per-process only — a DCS_server.exe restart resets all
in-memory state to OOB (there is no persistence yet).
5. Observe (dcs.log)
Mission-script logging goes through env.info() and lands in:
Saved Games/DCS/Logs/dcs.log
Read that file (or the log-reading MCP tool the DCS Studio server exposes, if present) to confirm
your change fired: init lines, scheduler ticks, reserve/strength/phase logs. What each system
should print and roughly when is the live-validation skill's observation checklist.
Guardrails baked into this loop
Verified type names only — query the live DB, never guess
Before you use any new unit type name or weapon CLSID in code, query the live DCS DB through
the DCS Studio bridge/MCP and confirm the exact string. The DB is reachable even at the main menu.
Guessing a type name silently produces coalition.addGroup failures or wrong loadouts. This is how
the verified statics (.Ammunition depot, Fuel tank, Comms tower M, .Command Center) and all
pylon CLSIDs were established — follow the same path for anything new.
Spawners must refund the reserve pool on every failure branch
Every spawn passes through the finite per-side reserve (supply.consume_side on spawn,
supply.recycle_side on RTB/failure). If a spawn consumes reserve and then bails — most commonly
coalition.addGroup returning nil, or a "no friendly base" early return — you must
recycle_side the exact amount back before returning. A missed refund silently drains a pool
forever (this class of bug produced 3 HIGH findings in the Cluster A review). Audit every new
early-return and error branch in any spawner code path for the matching refund. Wrap addGroup in
pcall where a throw is possible and refund in the failure arm.
