Skip to content

Codex Pro Dispatch v1.2.0

Choose a tag to compare

View all tags
@iannwu iannwu released this 04 Sep 08:54
v1.2.0
0778e66
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Codex Pro Dispatch v1.2.0 adds lean, bounded delivery for long ChatGPT Pro results while preserving the existing at-most-one-send safety model.

Highlights:

  • Continues long results through the same confirmed Pro conversation in bounded chunks.
  • Reassembles accepted UTF-8 payload bytes exactly in one private transient file.
  • Treats 10,000 bytes as generation guidance, not an acceptance gate.
  • Stops on malformed, truncated, out-of-sequence, or over-16-chunk responses without automatic resend.

Requirements:

  • Official ChatGPT/Codex desktop app on macOS
  • A dedicated ChatGPT conversation with Pro selected
  • Current Codex CLI with plugin marketplace support
  • Python 3.9 or newer
  • GitHub connector only for repository-write assignments

Verification:

  • 77 local tests passed
  • macOS and Ubuntu CI passed on the merge commit
  • Native desktop acceptance reconstructed 42,620 bytes across five chunks with exact SHA-256 equality
  • An intact 10,291-byte chunk was accepted under the advisory 10K target

See README.md and docs/releases/v1.2.0-long-result-acceptance.md for installation, limits, and redacted acceptance evidence.

Assets 2
Loading