Fixes from the rest of the live session #122

2026-10-08T02:17:00Z

dev-agent commented

2026-10-08 02:17:00 +00:00

The rest of what the live session of the new loop found (after #119):

Near-miss commands. In an item's thread, abort, cancel, drop it, quit and similar get "Did you mean abandon? …". Nothing changes, and the driver isn't called.
The driver claiming a stop. The system prompt now says only the operator stops or closes an item, so the driver must never say it has done either.
Reassigning an abandoned issue. If an issue is assigned to the agent again after its item closed, a new item starts. It records the item before it (follows) and says so in its opening fact. gitea.assigned_at reads the latest assignment from the issue's timeline, paged to the end. item_for_issue returns the latest item. The walk's one_item_per_issue now counts open items only. D1 in Architecture.md says so too.
Abandoning a blocked issue. When Gitea won't close the issue because of open dependencies, the message says so and names them.
No restate once agreed. An item that has started isn't offered restate (the live run restated a filed item at 01:38).

Each fix has a test in test_loop_live_fixes.py. assigned_at has its own test in test_gitea.py.

Local runs: unit 744 passed, integration 21 passed (7 opt-in skips), smoke e2e 3 passed.

🤖 Generated with Claude Code

https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE

The rest of what the live session of the new loop found (after #119): - **Near-miss commands.** In an item's thread, `abort`, `cancel`, `drop it`, `quit` and similar get "Did you mean `abandon`? …". Nothing changes, and the driver isn't called. - **The driver claiming a stop.** The system prompt now says only the operator stops or closes an item, so the driver must never say it has done either. - **Reassigning an abandoned issue.** If an issue is assigned to the agent again after its item closed, a new item starts. It records the item before it (`follows`) and says so in its opening fact. `gitea.assigned_at` reads the latest assignment from the issue's timeline, paged to the end. `item_for_issue` returns the latest item. The walk's `one_item_per_issue` now counts open items only. D1 in Architecture.md says so too. - **Abandoning a blocked issue.** When Gitea won't close the issue because of open dependencies, the message says so and names them. - **No `restate` once agreed.** An item that has started isn't offered `restate` (the live run restated a filed item at 01:38). Each fix has a test in `test_loop_live_fixes.py`. `assigned_at` has its own test in `test_gitea.py`. Local runs: unit 744 passed, integration 21 passed (7 opt-in skips), smoke e2e 3 passed. 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE

dev-agent added 1 commit 2026-10-08 02:17:00 +00:00

Fixes from the rest of the live session

CI / checks (push) Skipped

CI / checks (pull_request) Successful in 2m13s

Details

CI / e2e (pull_request) Successful in 3m22s

Details

0064ec7696

- A near miss for abandon ("abort", "cancel", "drop it", ...) in an item's
  thread is asked about, and changes nothing.
- The driver is told only the operator stops or closes an item, so it
  doesn't claim to have done either.
- An issue assigned again after its item closed starts a new item that
  notes the one before (Gitea's timeline gives the assignment time);
  item_for_issue returns the latest item.
- Abandoning an issue Gitea won't close because of open dependencies says
  so, naming them.
- An agreed item isn't offered restate.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE

dongar approved these changes

2026-10-08 02:36:47 +00:00

dongar merged commit 09232dcbdd into main

2026-10-08 02:36:52 +00:00

dongar deleted branch live-findings

2026-10-08 02:36:52 +00:00

Sign in to join this conversation.

2 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: dev/nisse#122