Conversation
The Agent tool exposes `resume` and `get_subagent_result`, both keyed by the subagent session ID, but a foreground run that completes returns only the subagent's report text; the ID is only in `details`, which the model never sees. Background runs already put the ID in text. Prefix the completed text with `Subagent <id> completed.` so the model can pass it back, matching the other status branches.
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Problem
The
Agenttool exposesresume("Existing subagent session ID to continue instead of creating a new session") andget_subagent_result(agent_id), both keyed by the subagent session ID. But when a foreground subagent completes, the tool result text is only the subagent's report; the session ID is present only indetails, which the model never sees. Background runs already includeSession ID: …in their text, and every other status branch (running,aborted,interrupted,failed) includes the ID too.In practice this makes
resumeunreachable for foreground runs. Observed with a builder profile that hasmax_turns: it hit the limit and handed back a partial report, the parent model tried toresumeas instructed, guessed three wrong IDs (Subagent not found), and had to start a fresh session and redo the work.Change
subagentFinalTextnow prefixes the completed text withSubagent <id> completed.(andSubagent <id> completed without text output.when the run produced no text), matching the format of the other status branches. The report body is unchanged.detailsis untouched.This also applies to the background completion notification delivered via
notifyParent, so both paths now consistently carry the ID.Verification
lib/subagent-extension.test.mjs: foreground completion,get_subagent_result(with and without text), andresumenow assert the ID-prefixed text.lib/subagent-runtime.test.mjs: background completion notification asserts the ID-prefixed text.npm test(38 in the touched suites; full suite 1094 on the fork branch),tsc --noEmitclean.