fix(task): report error to parent when subagent finishes with error (#52378)

[JerryLiu369]

Issue for this PR

Closes #52378

Type of change

• Bug fix
• New feature
• Refactor / code improvement
• Documentation

What does this PR do?

When a subagent assistant message ends with finish: "error" without an explicit error object, TaskTool.runTask previously ignored the error finish status and fell through to returning empty text parts, reporting the subagent task as successfully completed (state="completed").

This fix ensures TaskTool.runTask checks for finish === "error" on the assistant message and fails the task accordingly, surfacing the failure to the parent task runner with state="error". Also adds a unit test in task.test.ts.

How did you verify your code works?

• Tested with bun test test/tool/task.test.ts in packages/opencode. All 23 tests pass cleanly.

Screenshots / recordings

N/A

Checklist

• I have read the CONTRIBUTING document.
• My code follows the code style of this project.
• I have added tests to cover my changes.
• All new and existing tests passed.

[github-actions]

The following comment was made by an LLM, it may be inaccurate:

[r266-tech]

The explicit finish: "error" check addresses the old task-tool path, but this PR targets dev/packages/opencode. The current contribution guide asks for v2, where subagent jobs run through packages/core instead.

On current V2 55dde881, the successfully ended stream synthesizes an assistant failure only for finish === "unknown"; "error" still does not enter that branch. The V2 reproduction already shared in #52378 exercises the real runner, job and database for both foreground and background tasks. The stubbed TaskPromptOps test here does not cover that path. The issue also has a separately reproduced length case; it needs an explicit scope decision rather than being treated as covered by the error-only check.

Would you be able to port the fix to the V2 owner and reuse packages/core/test/tool-subagent.test.ts for both delivery modes? This comment is based on the current source comparison and the linked earlier synthetic event reproduction, not a new live-provider test. AI-assisted review; no competing PR opened.

[JerryLiu369]

@r266-tech Ported to the V2 owner path as requested, pushed to this branch:

• packages/core/src/session/runner/llm.ts now synthesizes an assistant failure (InvalidProviderOutputReason) when a successfully ended stream records finish unknown or error, and suppresses Step.Ended/continuation after that failure so it surfaces as a terminal step instead of silent success.
• packages/core/src/session/runner/publish-llm-event.ts exposes hasAssistantFailure used for the suppression.
• Coverage: new SessionRunnerLLM case projects an error finish as a terminal assistant step failure (foreground path), plus a background delivery regression in packages/opencode/test/tool/task.test.ts asserting the parent job ends in error state with the task_id message. The original foreground task test is untouched.
• Scope decision on length: explicitly left out, with a code comment — length is a valid truncation outcome that may carry usable partial output, so it stays outside this subagent-failure fix.

Verification on this box: packages/opencode task suite 24/24 green, packages/core session-runner suite 88/88 green, scoped tsgo typecheck clean for both packages. Repo-wide bun typecheck cannot complete in this environment (tsgo exceeds 300s and gets SIGTERM), so that one is left to CI.

[thdxr]

This branch has been upgraded to V2. We're no longer taking PRs for V1. If you think this is still relevant, please port it over and open it for V2.

— from 𝕺𝖕𝖊𝖓𝕮𝖔𝖉𝖊