codex's queue as the message-injection door for codex panes

What it does

Delivers messages to a codex pane through codex's own queue instead of pasting them into its composer.

Source

ganja-code-tvh (adopt codex's queue as the native injection door for codex panes).

The bead is closed when this issue is published, with its external reference pointing here (#38).

File and line references in this issue are to main as of 75a7a92.

Against the three criteria

• Direction (README): The README names teammates driven through other vendors' CLIs, codex among them (README.md:6-8).
• Today's terminal users: Codex panes receive messages through tmux bracketed paste today, one message at a time per member (crates/ganja-teammate-local/src/shim_tui.rs:44-56). No defect is recorded against that door.
• Evidence: No code touches a codex queue, and nothing records whether that queue is a stable door. The prerequisite the bead names — finding the session under ~/.codex/sessions — exists since D515 for readback (crates/ganja-teammate-local/src/codex.rs:501-512).

Depends on

No other idea. A probe of codex's queue comes first and is part of the work.

Verdict

• Verdict: do
• Size: M — one plan and one decision number.
• Exception to the probe rule: yes. #36 says an idea that needs a measurement or a probe first is not do; this one is an explicit exception: the probe of codex's queue is part of the work.
• Date: 2026-09-19
• Verdict ticket: #40

Theme: #53

Start with crates/ganja-teammate-local/src/shim_tui.rs:44-56 and crates/ganja-teammate-local/src/codex.rs:501-512, then probe Codex's queue as described by the issue. Done means establishing a viable native queue injection door and delivering messages through it instead of tmux bracketed paste.