CAT manual vs implementation — gap audits

Per-radio comparison of each transceiver's official CAT/CI-V reference manual against what rigplane-core actually implements, across three surfaces:

1. Backend — does a method exist (src/rigplane/backends/<driver>/)?
2. Profile — does rigs/<radio>.toml declare the capability + command strings?
3. Validation — does src/rigplane/validation/registry/ have a round-trip (RMVR) check, and what is its live status?

Each audit produces a 5-column command matrix (CAT cmd · Documented · Backend · Profile · Validation+live) plus four gap lists:

• A. Under-declared — backend implements it, but profile/registry can't see it.
• B. Validation gaps — implemented + declared, but no round-trip check (presence-only or nothing).
• C. Missing backend — documented operator command, no backend method at all.
• D. Mismatch / wrong — declared/checked but behaves differently, or a live FAIL tracing to a real mismatch.

Each gap row cross-references an existing or NEW Linear ticket (team RigPlane, key MOR).

Audits

Radio	Driver	Manual	Status
FTX-1 (Yaesu)	yaesu_cat	CAT OM v2507-B	✅ ftx1.md — done, live-validated
IC-7610 (Icom)	icom_civ	CI-V Reference Guide rev 1a	✅ ic7610.md — done, live-validated (3 scope FAILs → MOR-664)
IC-7300 (Icom)	icom_civ (shared)	Full Manual v6 §19	✅ ic7300.md — done, static/template (no live run)
Xiegu X6200	ic705 serial (CI-V-like)	Radioddity CI-V V1.0.6	✅ x6200.md — done, doc-vs-code (live not run)
Discovery TX-500 (lab599)	none (Kenwood-CAT, unimplemented)	Lab599 CAT Protocol rev.2	✅ tx500.md — done, doc-vs-code (no backend, no hw)

Source manuals

Downloaded vendor PDFs and their extracted .txt are kept locally under manuals/, which is git-ignored — these are copyrighted Icom / Yaesu / Xiegu / lab599 documents and this is a public repo, so we do not redistribute them. Each audit .md cites the exact manual revision + source URL so anyone can re-fetch it. Command mnemonics / opcodes referenced in the audits are interface facts, not reproductions of the manuals.