fix(b53d-00b0): replace chained ticket comment+transition with single --reason flag (merge worktree-20260327-075020)

Author: JoeOakhartNava

plugins/dso/skills/debug-everything/SKILL.md

@@ -436,7 +436,7 @@ Sub-agent prompt: Read `$PLUGIN_ROOT/skills/debug-everything/prompts/auto-fix.md
 ### Orchestrator Actions After Sub-Agent Returns
 
 1. Verify the sub-agent's report
-2. Close any issues resolved by auto-fix: `.claude/scripts/dso ticket comment <id> "Resolved by auto-fix (format/lint)"` then `.claude/scripts/dso ticket transition <id> open closed`
+2. Close any issues resolved by auto-fix: `.claude/scripts/dso ticket transition <id> open closed --reason="Fixed: resolved by auto-fix (format/lint)"`
 3. Update the failure inventory with remaining errors
 4. **CONTEXT ANCHOR**: After the commit workflow completes, continue immediately at Step 5 below (Phase 4). Do NOT stop or wait for user input after committing.
 
@@ -676,7 +676,7 @@ Sub-agent prompt: Read `$PLUGIN_ROOT/skills/debug-everything/prompts/post-batch-
 
 | Sub-agent outcome | Action |
 |------------------|--------|
-| Success + tests pass | `.claude/scripts/dso ticket comment <id> "Fixed: <summary>"` then `.claude/scripts/dso ticket transition <id> open closed` |
+| Success + tests pass | `.claude/scripts/dso ticket transition <id> open closed --reason="Fixed: <summary>"` |
 | Partial success | `.claude/scripts/dso ticket comment <id> "Partial: <details>."` |
 | Failure | `.claude/scripts/dso ticket transition <id> open` then `.claude/scripts/dso ticket comment <id> "Failed: <error>."` |
 | Regression | Revert changes (`git checkout -- <files>`), reopen, note regression |

plugins/dso/skills/sprint/SKILL.md

@@ -161,7 +161,7 @@ This ensures `issue-quality-check.sh` passes and sub-agents can self-serve their
 1. Run `.claude/scripts/dso ticket list` and filter for in-progress tasks under `<epic-id>` for interrupted tasks
 2. For each in-progress task, run `.claude/scripts/dso ticket show <id>` and parse its notes for CHECKPOINT lines
 3. Apply checkpoint resume rules:
-   - **CHECKPOINT 6/6 ✓** — task is fully done; fast-close: verify files exist, then `.claude/scripts/dso ticket comment <id> "Fixed: <summary>"` + `.claude/scripts/dso ticket transition <id> open closed`
+   - **CHECKPOINT 6/6 ✓** — task is fully done; fast-close: verify files exist, then `.claude/scripts/dso ticket transition <id> open closed --reason="Fixed: <summary>"`
    - **CHECKPOINT 5/6 ✓** — near-complete; fast-close: spot-check files and close without re-execution
    - **CHECKPOINT 3/6 ✓ or 4/6 ✓** — partial progress; re-dispatch with resume context: include the highest checkpoint note in the sub-agent prompt so it can continue from that substep
    - **CHECKPOINT 1/6 ✓ or 2/6 ✓** — early progress only; revert to open with `.claude/scripts/dso ticket transition <id> open` for full re-execution

tests/skills/test-no-chained-ticket-close.sh

@@ -0,0 +1,64 @@
+#!/usr/bin/env bash
+# tests/skills/test-no-chained-ticket-close.sh
+# Validates that agent-facing docs do NOT chain ticket comment + ticket
+# transition in a way that suggests a single Bash tool call.
+#
+# Bug: b53d-00b0 — chained commands exceed ~73s timeout, silently dropping
+# the transition.
+#
+# Usage: bash tests/skills/test-no-chained-ticket-close.sh
+
+set -uo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+REPO_ROOT="$(cd "$SCRIPT_DIR/../.." && pwd)"
+
+source "$REPO_ROOT/tests/lib/assert.sh"
+
+# =============================================================================
+# Test 1: sprint SKILL.md does not chain comment + transition
+# =============================================================================
+echo ""
+echo "--- test_sprint_no_chained_comment_transition ---"
+_snapshot_fail
+
+_SPRINT_CHAIN="not_found"
+grep -q 'ticket comment.*+.*ticket transition\|ticket comment.*then.*ticket transition' \
+    "$REPO_ROOT/plugins/dso/skills/sprint/SKILL.md" 2>/dev/null && _SPRINT_CHAIN="found"
+assert_eq "test_sprint_no_chained_comment_transition" "not_found" "$_SPRINT_CHAIN"
+
+assert_pass_if_clean "test_sprint_no_chained_comment_transition"
+
+# =============================================================================
+# Test 2: debug-everything SKILL.md does not chain comment + transition
+# =============================================================================
+echo ""
+echo "--- test_debug_everything_no_chained_comment_transition ---"
+_snapshot_fail
+
+_DEBUG_CHAIN="not_found"
+grep -q 'ticket comment.*then.*ticket transition\|ticket comment.*+.*ticket transition' \
+    "$REPO_ROOT/plugins/dso/skills/debug-everything/SKILL.md" 2>/dev/null && _DEBUG_CHAIN="found"
+assert_eq "test_debug_everything_no_chained_comment_transition" "not_found" "$_DEBUG_CHAIN"
+
+assert_pass_if_clean "test_debug_everything_no_chained_comment_transition"
+
+# =============================================================================
+# Test 3: bug-type close commands use --reason (not separate comment)
+# Sprint and debug-everything should use --reason for bug close, not a
+# separate ticket comment call.
+# =============================================================================
+echo ""
+echo "--- test_bug_close_uses_reason_flag_in_sprint ---"
+_snapshot_fail
+
+# Sprint SKILL.md line ~164 (CHECKPOINT 6/6) should use --reason
+_SPRINT_REASON="not_found"
+grep -q 'ticket transition.*closed.*--reason\|--reason.*Fixed' \
+    "$REPO_ROOT/plugins/dso/skills/sprint/SKILL.md" 2>/dev/null && _SPRINT_REASON="found"
+assert_eq "test_bug_close_uses_reason_flag_in_sprint" "found" "$_SPRINT_REASON"
+
+assert_pass_if_clean "test_bug_close_uses_reason_flag_in_sprint"
+
+# =============================================================================
+print_summary