Skip to content

Add model and effort flags to review commands#477

Closed
axisrow wants to merge 1 commit into
openai:mainfrom
axisrow:codex-review-flags
Closed

Add model and effort flags to review commands#477
axisrow wants to merge 1 commit into
openai:mainfrom
axisrow:codex-review-flags

Conversation

@axisrow

@axisrow axisrow commented Jul 11, 2026

Copy link
Copy Markdown

Propagate --model and --effort to /codex:review and /codex:adversarial-review, matching the existing task behavior.

Validation:

  • node --test tests/commands.test.mjs
  • node --test tests/runtime.test.mjs

@axisrow

axisrow commented Jul 11, 2026

Copy link
Copy Markdown
Author

Closing this as the wrong fix.

The approach here — forwarding model/effort into the review/start RPC params — doesn't actually make the flags work. Probing the real app-server (codex-cli 0.144.1), review/start accepts model/effort without error but the Rust side drops unknown fields (serde, no deny_unknown_fields), and ReviewStartParams has no model/effort field in the first place (TS2353 on a direct type-probe). The model field is inert on the server across thread/start/turn/start/review/start — the effective override is -c model="…" at spawn time (#408), not an RPC field.

For effort: adversarial-review can be fixed via turn/start (its params include effort) per #476; native review has no effort field in the protocol, so it's not fixable on the plugin side without the #408-style spawn override.

Left detailed notes + the server-probe evidence on #476 and #408. Folding into those instead of keeping a separate incompatible change open.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant