codex-pool preset revision
Use this child for the named built-in codex-pool preset. codexPoolPreset in tui/internal/preset/preset.go:1476 mirrors codex: gpt-5.6-sol, https://chatgpt.com/backend-api/codex, empty api_key_env, thinking xhigh, web_search, and vision. Pooling changes OAuth account selection, not the model or endpoint. Keep provider codex-pool distinct from codex.
Template-specific settings
Use the official Codex models and authentication pages, then verify served acceptance against the same Codex route for each relevant pooled account. Do not use the standard OpenAI API catalog. The pool is not a model catalog. Do not turn one account's observation into a claim about every account.
TUI surfaces to revise
Start at codexPoolPreset in tui/internal/preset/preset.go. Keep providerModels["codex-pool"] exactly aligned with providerModels["codex"], including gpt-5.6-sol, gpt-6-astra, gpt-5.6-terra, and gpt-5.6-luna in that default-first order, with matching modelHasVision entries and the Codex four-level codexThinkingOptions in tui/internal/tui/preset_editor.go. Recheck the constructor's vision, /codex endpoint, OAuth-only fields, and the pool store only when pool selection or file behavior is part of the change. Follow tui/internal/tui/SKILL.md and tui/CONTRACT.md.
Pool data, selection, and edits
The pool file is $LINGTAI_TUI_DIR/codex-auth-pool.json, or
~/.lingtai-tui/codex-auth-pool.json when that variable is unset. The pool
stores only refs and integer weights.
The flat shape is {"version": 1, "accounts": [{"path": "...", "weight": 1}]};
the classified shape is {"version": 2, "models": {"<exact model>": [...]}}.
Presence of the models key is what classifies the pool, not its size or
version. Classified pools refuse flat TUI edits with
errCodexPoolModelClassified. Invalid or missing pool data, or an unmatched exact model, falls back to the legacy single-account auth path. Weight 0 means the account is present but disabled. Selection happens at adapter/service construction, is weighted and sticky within one agent wake/session, and does not reselect an already-running session.
Configured weights are inputs, not measured shares, and the pool excludes molt_count.
Manual edits require Exact authorization; Timestamped backup; Exact-old-value or hash gate; same-directory atomic rename; validation with the live
load_codex_auth_pool; and preservation on failure. Preserve the original file on any validation failure. Display labels/email only; never token contents or absolute auth paths (the exact safety rule is: Never print token/auth contents or absolute auth paths). Source anchors:
tui/internal/tui/codex_pool_store.go:11-330,
login.go:171-201,285-299,603-702, auth/codex_pool.py:72-323, and
_register.py:424-497.
To inspect live OAuth quota for one pooled account, complete initialize, send
account/rateLimits/read with structurally null params, and optionally watch
account/rateLimits/updated. This is per-account and reports usedPercent,
windowDurationMins, and resetsAt; it is not token or credential data.
Reviewed deterministic revision
Prepare an evidence-bound manifest and explicit input, then run lingtai-tui presets revise --manifest PATH --input PATH --mode dry-run|check|apply [--output-dir PATH]. Review the JSON plan, use dry-run/check before apply, and apply only to a new explicit output directory. revision.go validates hashes, route bindings, and evidence and preserves unowned bytes. This amendment does update the picker and vision metadata while keeping the constructor/default at gpt-5.6-sol behind the exact-route availability gate; pool data is unchanged.
Maintenance: If the relevant TUI preset/page is revised, check whether this sub-skill also needs revision and, if so, include it in the same PR.
Operations
For save, endpoint/capability, availability, activation/refresh, or
troubleshooting, use the five shared operation children under
reference/operations/; use
reference/operations/endpoint-capabilities/SKILL.md for per-account OAuth
quota. This child owns Codex-pool selection and edit facts.