fix: half-clone counts progress messages as user messages

The grep chains matched "type":"user" anywhere in the raw JSON line,
including inside nested subagent progress data. This inflated the clean
user message count (e.g. 53 vs actual 28), causing the cut point to
land on a progress message instead of a real user turn.

Replace all grep chains with a single filter_clean_user_msgs helper
that extracts the top-level type (first "type":"..." in the line).
Also filters isMeta, interrupted, and tool_result in one pass.

Fix 4 pre-existing test failures (off-by-one from uncounted synthetic
marker message, double-tag test checking wrong line). Add test for
progress messages with nested type:user.

Bump plugin to 0.14.9.

Author: ykdojo

.claude-plugin/marketplace.json

@@ -9,7 +9,7 @@
       "name": "dx",
       "source": "./",
       "description": "Developer experience essentials: GitHub Actions debugging, conversation cloning/half-cloning, context handoffs, and Reddit research via Gemini CLI",
-      "version": "0.14.8"
+      "version": "0.14.9"
     }
   ]
 }

.claude-plugin/plugin.json

@@ -1,7 +1,7 @@
 {
   "name": "dx",
   "description": "Developer experience essentials: GitHub Actions debugging, conversation cloning/half-cloning, context handoffs, and Reddit research via Gemini CLI",
-  "version": "0.14.8",
+  "version": "0.14.9",
   "author": {
     "name": "YK",
     "email": "yoyoyosss@wearehackerone.com"

scripts/half-clone-conversation.sh

@@ -99,6 +99,32 @@ get_project_from_conv_file() {
     echo "$project_dirname" | sed 's|^-|/|' | sed 's|-|/|g'
 }
 
+# Filter JSONL to clean user messages only.
+# Checks the FIRST "type":"..." in each line (the top-level type), so nested
+# "type":"user" inside progress/subagent data is ignored.
+# Also excludes tool_results, isMeta skill expansions, and interrupted messages.
+# Usage: filter_clean_user_msgs < file.jsonl         (output matching lines)
+#        filter_clean_user_msgs -n < file.jsonl       (with line numbers: NR:line)
+#        filter_clean_user_msgs -c < file.jsonl       (count only)
+filter_clean_user_msgs() {
+    local mode="lines"
+    if [ "${1:-}" = "-n" ]; then mode="numbered"; fi
+    if [ "${1:-}" = "-c" ]; then mode="count"; fi
+    awk -v mode="$mode" '
+        match($0, /"type":"[^"]*"/) {
+            t = substr($0, RSTART+8, RLENGTH-9)
+            if ((t == "user" || t == "queue-operation") &&
+                index($0, "\"type\":\"tool_result\"") == 0 &&
+                index($0, "\"isMeta\":true") == 0 &&
+                index($0, "Request interrupted by user") == 0) {
+                count++
+                if (mode == "numbered") print NR":"$0
+                else if (mode == "lines") print
+            }
+        }
+        END { if (mode == "count") print count+0 }'
+}
+
 # Pre-generate UUIDs for the awk script
 pre_generate_uuids() {
     local count="$1"
@@ -136,11 +162,11 @@ preview_conversation() {
     local total_lines
     total_lines=$(wc -l < "$source_file" | tr -d ' ')
     local first_user_text
-    first_user_text=$(grep -E '"type":"(user|queue-operation)"' "$source_file" | grep -v '"type":"tool_result"' | head -1 | \
+    first_user_text=$(filter_clean_user_msgs < "$source_file" | head -1 | \
         grep -oE '"(content|text)":"[^"]*"' | head -1 | \
         LC_ALL=C sed 's/"content":"//;s/"text":"//;s/"$//' | cut -c1-120 || true)
     local last_user_text
-    last_user_text=$(grep -E '"type":"(user|queue-operation)"' "$source_file" | grep -v '"type":"tool_result"' | tail -1 | \
+    last_user_text=$(filter_clean_user_msgs < "$source_file" | tail -1 | \
         grep -oE '"(content|text)":"[^"]*"' | head -1 | \
         LC_ALL=C sed 's/"content":"//;s/"text":"//;s/"$//' | cut -c1-120 || true)
 
@@ -185,7 +211,7 @@ half_clone_conversation() {
     # Count "clean" user messages (not tool_results - those require a preceding tool_use)
     # A clean user message is one where we can start a conversation
     local total_clean_user_messages
-    total_clean_user_messages=$(grep -E '"type":"(user|queue-operation)"' "$source_file" | grep -cv '"type":"tool_result"' || echo "0")
+    total_clean_user_messages=$(filter_clean_user_msgs -c < "$source_file")
     log_info "Total clean user messages in conversation: $total_clean_user_messages"
 
     if [ "$total_clean_user_messages" -lt 2 ]; then
@@ -200,9 +226,8 @@ half_clone_conversation() {
     keep_clean_count=$((total_clean_user_messages - skip_clean_count))
 
     # OPTIMIZED: Find the line number where the target clean user message starts
-    # Use grep -n to get all clean user message line numbers in one pass
     local clean_user_line_numbers
-    clean_user_line_numbers=$(grep -nE '"type":"(user|queue-operation)"' "$source_file" | grep -v '"type":"tool_result"' | cut -d: -f1)
+    clean_user_line_numbers=$(filter_clean_user_msgs -n < "$source_file" | cut -d: -f1)
 
     # Get the line number of the (skip_clean_count + 1)th clean user message
     local skip_count
@@ -232,10 +257,9 @@ half_clone_conversation() {
     local last_clone_cmd_line=0
     local last_clean_user_line=0
 
-    # Get all user message lines with line numbers (much faster than per-line grep)
-    # Filter to clean user messages (not tool_result, not isMeta)
+    # Get all clean user message lines with line numbers
     local clean_user_lines
-    clean_user_lines=$(grep -nE '"type":"(user|queue-operation)"' "$source_file" | grep -v '"type":"tool_result"' | grep -v '"isMeta":true' || true)
+    clean_user_lines=$(filter_clean_user_msgs -n < "$source_file" || true)
 
     if [ -n "$clean_user_lines" ]; then
         # Get the last clean user message line
@@ -356,6 +380,13 @@ half_clone_conversation() {
         return line
     }
 
+    function get_top_type(line) {
+        if (match(line, /"type":"[^"]*"/)) {
+            return substr(line, RSTART+8, RLENGTH-9)
+        }
+        return ""
+    }
+
     function halve_number(line, field,    pattern, num, halved) {
         pattern = "\"" field "\":[0-9]+"
         if (match(line, pattern)) {
@@ -416,11 +447,16 @@ half_clone_conversation() {
             gsub("\"messageId\":\"" old_msgid "\"", "\"messageId\":\"" new_msgid "\"", line)
         }
 
-        # Tag first user message (including queue-operation messages)
-        if (first_user && (index(line, "\"type\":\"user\"") > 0 || index(line, "\"type\":\"queue-operation\"") > 0)) {
-            gsub("\"content\":\"", "\"content\":\"" clone_tag " ", line)
-            gsub("\"text\":\"", "\"text\":\"" clone_tag " ", line)
-            first_user = 0
+        # Tag first genuine user message (check top-level type, skip isMeta/interrupted)
+        if (first_user) {
+            top_type = get_top_type(line)
+            if ((top_type == "user" || top_type == "queue-operation") &&
+                index(line, "\"isMeta\":true") == 0 &&
+                index(line, "Request interrupted by user") == 0) {
+                gsub("\"content\":\"", "\"content\":\"" clone_tag " ", line)
+                gsub("\"text\":\"", "\"text\":\"" clone_tag " ", line)
+                first_user = 0
+            }
         }
 
         # Halve token counts
@@ -460,16 +496,16 @@ half_clone_conversation() {
     # Update history.jsonl
     log_info "Updating history file..."
 
-    # Get display text from first user message in the KEPT portion
+    # Get display text from first clean user message in the KEPT portion
     local display_text
-    display_text=$(tail -n +"$((skip_count + 1))" "$source_file" | grep -E '"type":"(user|queue-operation)"' | head -1 | \
+    display_text=$(tail -n +"$((skip_count + 1))" "$source_file" | filter_clean_user_msgs | head -1 | \
         grep -oE '"content":"[^"]*"' | head -1 | \
         LC_ALL=C sed 's/"content":"//;s/"$//' | \
         head -c 200 || echo "[Half-cloned conversation]")
 
     if [ -z "$display_text" ]; then
         # Try array format
-        display_text=$(tail -n +"$((skip_count + 1))" "$source_file" | grep -E '"type":"(user|queue-operation)"' | head -1 | \
+        display_text=$(tail -n +"$((skip_count + 1))" "$source_file" | filter_clean_user_msgs | head -1 | \
             grep -oE '"text":"[^"]*"' | head -1 | \
             LC_ALL=C sed 's/"text":"//;s/"$//' | \
             head -c 200 || echo "[Half-cloned conversation]")

scripts/test-half-clone.sh

@@ -108,9 +108,9 @@ get_new_session_from_output() {
 }
 
 # Test 1: 6 messages (3 user, 3 assistant) -> 3 clean user msgs, skip 1, keep 2
-# Starts at user message 2 (line 3), keeps lines 3-6 = 4 messages + 1 reference = 5
+# Output: marker(1) + kept lines 3-6(4) + reference(1) = 6
 test_even_messages() {
-    log_test "6 messages (3 clean user): should keep 4 lines + 1 reference = 5"
+    log_test "6 messages (3 clean user): should produce 1 marker + 4 kept + 1 reference = 6"
 
     local session_id
     session_id=$(create_test_conversation 6)
@@ -128,17 +128,17 @@ test_even_messages() {
 
     local count
     count=$(count_messages "$new_file")
-    if [ "$count" -eq 5 ]; then
-        log_pass "Kept 5 messages (4 + reference)"
+    if [ "$count" -eq 6 ]; then
+        log_pass "Kept 6 messages (marker + 4 kept + reference)"
     else
-        log_fail "Expected 5 messages, got $count"
+        log_fail "Expected 6 messages, got $count"
     fi
 }
 
 # Test 2: 7 messages (4 user, 3 assistant) -> 4 clean user msgs, skip 2, keep 2
-# Starts at user message 3 (line 5), keeps lines 5-7 = 3 messages + 1 reference = 4
+# Output: marker(1) + kept lines 5-7(3) + reference(1) = 5
 test_odd_messages() {
-    log_test "7 messages (4 clean user): should keep 3 lines + 1 reference = 4"
+    log_test "7 messages (4 clean user): should produce 1 marker + 3 kept + 1 reference = 5"
 
     local session_id
     session_id=$(create_test_conversation 7)
@@ -151,17 +151,17 @@ test_odd_messages() {
 
     local count
     count=$(count_messages "$new_file")
-    if [ "$count" -eq 4 ]; then
-        log_pass "Kept 4 messages (3 + reference)"
+    if [ "$count" -eq 5 ]; then
+        log_pass "Kept 5 messages (marker + 3 kept + reference)"
     else
-        log_fail "Expected 4 messages, got $count"
+        log_fail "Expected 5 messages, got $count"
     fi
 }
 
 # Test 3: 4 messages (2 user, 2 assistant) -> 2 clean user msgs, skip 1, keep 1
-# Starts at user message 2 (line 3), keeps lines 3-4 = 2 messages + 1 reference = 3
+# Output: marker(1) + kept lines 3-4(2) + reference(1) = 4
 test_minimum_messages() {
-    log_test "4 messages (2 clean user): should keep 2 lines + 1 reference = 3"
+    log_test "4 messages (2 clean user): should produce 1 marker + 2 kept + 1 reference = 4"
 
     local session_id
     session_id=$(create_test_conversation 4)
@@ -174,10 +174,10 @@ test_minimum_messages() {
 
     local count
     count=$(count_messages "$new_file")
-    if [ "$count" -eq 3 ]; then
-        log_pass "Kept 3 messages (2 + reference)"
+    if [ "$count" -eq 4 ]; then
+        log_pass "Kept 4 messages (marker + 2 kept + reference)"
     else
-        log_fail "Expected 3 messages, got $count"
+        log_fail "Expected 4 messages, got $count"
     fi
 }
 
@@ -335,19 +335,21 @@ test_double_tagging() {
     new_session=$(get_new_session_from_output "$output")
     local new_file="${TEST_PROJECTS_DIR}/${TEST_PROJECT_DIRNAME}/${new_session}.jsonl"
 
-    # Should have two [HALF-CLONE ...] tags - the new one prepended to the existing one
-    local first_user_line
-    first_user_line=$(grep '"type":"user"' "$new_file" | head -1)
+    # The first user line is the synthetic marker ("Continued from session ...").
+    # The second user line is the first KEPT message from the original, which
+    # already had a [HALF-CLONE] tag and should now have a second one prepended.
+    local kept_user_line
+    kept_user_line=$(grep '"type":"user"' "$new_file" | grep -v "Continued from session" | head -1)
 
     # Count occurrences of [HALF-CLONE pattern
     local tag_count
-    tag_count=$(echo "$first_user_line" | grep -oE '\[HALF-CLONE [A-Z][a-z]+ [0-9]+ [0-9]+:[0-9]+\]' | wc -l | tr -d ' ')
+    tag_count=$(echo "$kept_user_line" | grep -oE '\[HALF-CLONE [A-Z][a-z]+ [0-9]+ [0-9]+:[0-9]+\]' | wc -l | tr -d ' ')
 
     if [ "$tag_count" -eq 2 ]; then
         log_pass "Double tagging works - found 2 [HALF-CLONE] tags"
     else
         log_fail "Expected 2 [HALF-CLONE] tags, found $tag_count"
-        echo "First user line: $first_user_line"
+        echo "Kept user line: $kept_user_line"
     fi
 }
 
@@ -397,6 +399,57 @@ test_thinking_blocks_stripped() {
     fi
 }
 
+# Test 11: Progress messages with nested "type":"user" should NOT count as clean user messages
+test_progress_with_nested_user_type() {
+    log_test "Progress messages with nested type:user should not count as clean user messages"
+
+    local session_id
+    session_id=$(uuidgen | tr '[:upper:]' '[:lower:]')
+    local conv_file="${TEST_PROJECTS_DIR}/${TEST_PROJECT_DIRNAME}/${session_id}.jsonl"
+
+    local uuid1 uuid2 uuid3 uuid4 uuid5 uuid6 uuid7 uuid8
+    uuid1=$(uuidgen | tr '[:upper:]' '[:lower:]')
+    uuid2=$(uuidgen | tr '[:upper:]' '[:lower:]')
+    uuid3=$(uuidgen | tr '[:upper:]' '[:lower:]')
+    uuid4=$(uuidgen | tr '[:upper:]' '[:lower:]')
+    uuid5=$(uuidgen | tr '[:upper:]' '[:lower:]')
+    uuid6=$(uuidgen | tr '[:upper:]' '[:lower:]')
+    uuid7=$(uuidgen | tr '[:upper:]' '[:lower:]')
+    uuid8=$(uuidgen | tr '[:upper:]' '[:lower:]')
+
+    # Conversation: U1, A1, progress(nested user), progress(nested user), U2, A2, U3 CORRECT, A3
+    # Real clean user messages: U1, U2, U3 = 3. Skip 1, keep 2. Start at U2.
+    # If progress counted: 5 "clean" msgs, skip 2, start at U2 (or worse, a progress line)
+    # With the fix: 3 clean msgs, skip 1, start at U2 -> first real content is U2
+    {
+        generate_message "$uuid1" "null" "$session_id" "user" "Question 1"
+        generate_message "$uuid2" "$uuid1" "$session_id" "assistant" "Answer 1"
+        # Progress messages with nested "type":"user" (simulates subagent Task progress)
+        echo "{\"type\":\"progress\",\"sessionId\":\"${session_id}\",\"data\":{\"message\":{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"type\":\"text\",\"text\":\"subagent input\"}]}}},\"uuid\":\"${uuid3}\",\"timestamp\":\"2025-01-01T00:00:00.000Z\"}"
+        echo "{\"type\":\"progress\",\"sessionId\":\"${session_id}\",\"data\":{\"message\":{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"type\":\"text\",\"text\":\"another subagent input\"}]}}},\"uuid\":\"${uuid4}\",\"timestamp\":\"2025-01-01T00:00:00.000Z\"}"
+        generate_message "$uuid5" "$uuid2" "$session_id" "user" "Question 2"
+        generate_message "$uuid6" "$uuid5" "$session_id" "assistant" "Answer 2"
+        generate_message "$uuid7" "$uuid6" "$session_id" "user" "Question 3 CORRECT START"
+        generate_message "$uuid8" "$uuid7" "$session_id" "assistant" "Answer 3"
+    } > "$conv_file"
+
+    local output
+    output=$(run_half_clone "$session_id")
+
+    local new_session
+    new_session=$(get_new_session_from_output "$output")
+    local new_file="${TEST_PROJECTS_DIR}/${TEST_PROJECT_DIRNAME}/${new_session}.jsonl"
+
+    # The output from the script should say "3" clean user messages
+    if echo "$output" | grep -q "Total clean user messages in conversation: 3"; then
+        log_pass "Progress messages with nested type:user correctly excluded from count (3 clean, not 5)"
+    else
+        local actual_count
+        actual_count=$(echo "$output" | grep "Total clean user messages" | grep -oE '[0-9]+')
+        log_fail "Expected 3 clean user messages, script counted $actual_count"
+    fi
+}
+
 # Main
 main() {
     echo "================================"
@@ -422,6 +475,7 @@ main() {
     test_history_entry
     test_double_tagging
     test_thinking_blocks_stripped
+    test_progress_with_nested_user_type
 
     echo ""
     echo "================================"