From acbbd8747eadbf2cc2d580e5de30aa5d00521bad Mon Sep 17 00:00:00 2001
From: Ian Gulliver <git@flamingcow.io>
Date: Fri, 16 Oct 2026 12:36:05 -0700
Subject: [PATCH] A start failure after admission leaves a session `running`
 with no turn

---
 TODO.md                    |  2 --
 internal/agents/manager.go | 39 +++++++++++++++++++++++++++++---------
 2 files changed, 30 insertions(+), 11 deletions(-)

diff --git a/TODO.md b/TODO.md
index 8517340..216bebf 100644
--- a/TODO.md
+++ b/TODO.md
@@ -8,8 +8,6 @@
   A `CommitAll` error is only logged (`internal/agents/manager.go:838–841`) and `FinishStage` still runs (`:885–889`), so `RecordStage` stores the old `git.Head` (`internal/agents/workflows.go:349–356`), the next stage branches without the turn's work, and `discardStages` later deletes the worktree that held it. On commit failure, treat the turn as unfinished: skip FinishStage, the handoff and the summary, and file the error in the task's inbox.
 - One bad session at startup stops byre from starting
   `Recover` returns on a `taskFor` or `message` error (`internal/agents/manager.go:300–303`, `:338–341`) and `server.New` treats it as fatal (`internal/server/server.go:72–76`), so one unreadable worktree keeps the service down. Handle a failed `message` like a failed `ensureLive` (logUndelivered, DropStageOutput, EndTurn with the error, discard and merge follow-up) and continue; a `taskFor` error skips only that session.
-- A start failure after admission leaves a session `running` with no turn
-  `launch` resets to idle only on `ensureLive`/`syncModel` errors (`internal/agents/manager.go:439–466`); a local `message` error with a live process (missing blob, failed `rev-parse HEAD`, failed event write) leaves it running, so later messages queue forever and Merge skips the task. When `message` fails before `send`, set the session idle with the error; a `send` failure is already handled by pump.
 - Codex mid-turn delivery deletes input before the steer is accepted
   Queued inputs are deleted or marked delivered and recorded as `user.message` (`internal/agents/queue.go:119`, `:158`; `internal/agents/inbox.go:404`) before `Steer`, whose error is only logged (`internal/agents/codex.go:167–174`). Return the consumed ids from `deliverMidTurn`; when Steer fails, re-insert the inputs in order and clear `delivered_at` so the turn-end Deliver sends them.
 - Inbox answer validation isn't atomic with resolving the question
diff --git a/internal/agents/manager.go b/internal/agents/manager.go
index 19e3f56..e005620 100644
--- a/internal/agents/manager.go
+++ b/internal/agents/manager.go
@@ -450,14 +450,17 @@ func (m *Manager) launch(ctx context.Context, sess api.AgentSession, task api.Ta
 		err = m.syncModel(ctx, l, sess.Model)
 	}
 	if err != nil {
-		serr := m.store.SetSessionState(ctx, sess.Id, api.AgentSessionStateIdle, actor, map[string]any{"error": err.Error()})
-		if serr != nil {
-			slog.Error("record failed start", "session", sess.Id, "err", serr)
-		}
+		m.failStart(ctx, sess.Id, actor, err)
 		return api.AgentSession{}, err
 	}
 
-	err = m.message(ctx, l, actor, text, images)
+	imgs, err := m.prepareMessage(ctx, l, actor, text, images)
+	if err != nil {
+		m.failStart(ctx, sess.Id, actor, err)
+		return api.AgentSession{}, err
+	}
+
+	err = l.sess.send(ctx, text, imgs)
 	if err != nil {
 		return api.AgentSession{}, err
 	}
@@ -465,6 +468,16 @@ func (m *Manager) launch(ctx context.Context, sess api.AgentSession, task api.Ta
 	return m.store.GetSession(ctx, sess.Id)
 }
 
+func (m *Manager) failStart(ctx context.Context, sessionID, actor string, err error) {
+	cctx, cancel := context.WithTimeout(context.WithoutCancel(ctx), 10*time.Second)
+	defer cancel()
+
+	serr := m.store.SetSessionState(cctx, sessionID, api.AgentSessionStateIdle, actor, map[string]any{"error": err.Error()})
+	if serr != nil {
+		slog.Error("record failed start", "session", sessionID, "err", serr)
+	}
+}
+
 func (m *Manager) syncModel(ctx context.Context, l *live, model string) error {
 	if model == l.model {
 		return nil
@@ -490,11 +503,19 @@ func transition(ctx context.Context, st *store.Store, id string, from, to api.Ag
 }
 
 func (m *Manager) message(ctx context.Context, l *live, actor, text string, images []string) error {
+	imgs, err := m.prepareMessage(ctx, l, actor, text, images)
+	if err != nil {
+		return err
+	}
+	return l.sess.send(ctx, text, imgs)
+}
+
+func (m *Manager) prepareMessage(ctx context.Context, l *live, actor, text string, images []string) ([]image, error) {
 	imgs := []image{}
 	for _, id := range images {
 		b, data, err := m.store.ReadBlob(ctx, id)
 		if err != nil {
-			return err
+			return nil, err
 		}
 		imgs = append(imgs, image{mediaType: b.ContentType, data: data})
 	}
@@ -503,7 +524,7 @@ func (m *Manager) message(ctx context.Context, l *live, actor, text string, imag
 	if l.implementer() {
 		head, err := git.Head(l.dir)
 		if err != nil {
-			return err
+			return nil, err
 		}
 		payload["head"] = head
 	}
@@ -519,7 +540,7 @@ func (m *Manager) message(ctx context.Context, l *live, actor, text string, imag
 	}
 	err := m.emit(ctx, l, actor, "user.message", payload)
 	if err != nil {
-		return err
+		return nil, err
 	}
 
 	m.mu.Lock()
@@ -532,7 +553,7 @@ func (m *Manager) message(ctx context.Context, l *live, actor, text string, imag
 	l.lastText = ""
 	m.mu.Unlock()
 
-	return l.sess.send(ctx, text, imgs)
+	return imgs, nil
 }
 
 func (m *Manager) ensureLive(ctx context.Context, sess api.AgentSession, task api.Task) (*live, error) {