- codex-run.sh review: CLI forbids scope flags + prompt (stdin '-' counts
as [PROMPT]); translate scope into prompt text, native review only when
prompt.md is empty. Add --skip-git-repo-check for consistency.
- agents/codex.md: marker contract (=== TASK FOR CODEX === / === END TASK ===,
legacy RU markers accepted) so wrapper instructions never leak into the
codex prompt and the wrapper never mistakes the task for its own job;
self-check against doing the task itself; preamble now states codex is a
subagent executor and process flows belong to the orchestrator; scope
flags as separate args; no nohup; Read before rewriting prompt.md.
- skills: wrapper framing line + marker block in prompt templates.
- All instructions and README translated to English (RU trigger phrases
kept in skill descriptions).
- tests: 3 new review-mode cases, ALL PASS.
- bump 0.1.2.
Follow-up to the previous fix: the stdin branch dropped "$@", so
options like -s read-only silently never reached codex. Stub now
records argv and tests assert flags + stdin prompt arrive together;
verified live (-s read-only => codex reports READONLY).
Options like -s read-only routed execution into the /dev/null branch,
so codex got no prompt at all. codex natively appends piped stdin as
a <stdin> block, so feeding prompt.md unconditionally is safe.
Background-task re-invoke after SendMessage resume proved unreliable:
subagent ended its turn with 'still running' and was never woken.
Agent now polls for the status file in a foreground call (9.5-min
chunks, repeated), so the result always lands in its final message.
admin