diff --git a/planner/planner-agent.sh b/planner/planner-agent.sh
index 24acdd9..1252c6a 100755
--- a/planner/planner-agent.sh
+++ b/planner/planner-agent.sh
@@ -190,6 +190,27 @@ OPEN_SUMMARY=$(echo "$OPEN_ISSUES" | jq -r '.[] | "#\(.number) [\(.labels | map(
 # Fetch vision-labeled issues specifically
 VISION_ISSUES=$(echo "$OPEN_ISSUES" | jq -r '.[] | select(.labels | map(.name) | index("vision")) | "#\(.number) \(.title)\n\(.body)"' 2>/dev/null || true)
 
+# Read supervisor metrics for trend analysis (last 7 days)
+METRICS_FILE="${FACTORY_ROOT}/metrics/supervisor-metrics.jsonl"
+METRICS_SUMMARY="(no metrics data — supervisor has not yet written metrics)"
+if [ -f "$METRICS_FILE" ] && [ -s "$METRICS_FILE" ]; then
+  _METRICS_CUTOFF=$(date -u -d '7 days ago' +%Y-%m-%dT%H:%M)
+  METRICS_SUMMARY=$(awk -F'"' -v cutoff="$_METRICS_CUTOFF" \
+    'NF >= 4 && $2 == "ts" && $4 >= cutoff' "$METRICS_FILE" 2>/dev/null | \
+    jq -rs '
+      ( [.[] | select(.type=="ci") | .duration_min] | if length>0 then add/length|round else null end ) as $ci_avg |
+      ( [.[] | select(.type=="ci") | select(.status=="success")] | length ) as $ci_ok |
+      ( [.[] | select(.type=="ci")] | length ) as $ci_n |
+      ( [.[] | select(.type=="infra") | .ram_used_pct] | if length>0 then add/length|round else null end ) as $ram_avg |
+      ( [.[] | select(.type=="infra") | .disk_used_pct] | if length>0 then add/length|round else null end ) as $disk_avg |
+      ( [.[] | select(.type=="dev")] | last ) as $dev_last |
+      "CI (\($ci_n) pipelines): avg \(if $ci_avg then "\($ci_avg)min" else "n/a" end), success rate \(if $ci_n > 0 then "\($ci_ok * 100 / $ci_n | round)%" else "n/a" end)\n" +
+      "Infra: avg RAM \(if $ram_avg then "\($ram_avg)%" else "n/a" end) used, avg disk \(if $disk_avg then "\($disk_avg)%" else "n/a" end) used\n" +
+      "Dev (latest): \(if $dev_last then "\($dev_last.issues_in_backlog) in backlog, \($dev_last.issues_blocked) blocked (\(if $dev_last.issues_in_backlog > 0 then $dev_last.issues_blocked * 100 / $dev_last.issues_in_backlog | round else 0 end)% blocked), \($dev_last.pr_open) open PRs" else "n/a" end)
+    ' 2>/dev/null) || METRICS_SUMMARY="(metrics parse error)"
+  log "Metrics: ${METRICS_SUMMARY:0:120}"
+fi
+
 PHASE2_PROMPT="You are the planner for ${CODEBERG_REPO}. Your job: find gaps between the project vision and current reality.
 
 ## VISION.md (human-maintained goals)
@@ -204,6 +225,9 @@ ${VISION_ISSUES:-"(none)"}
 ## All open issues
 ${OPEN_SUMMARY}
 
+## Operational metrics (last 7 days from supervisor)
+${METRICS_SUMMARY}
+
 ## Task
 Identify gaps — things implied by VISION.md that are neither reflected in the project state nor covered by an existing open issue.
 
@@ -218,6 +242,7 @@ For each gap, output a JSON object (one per line, no array wrapper):
 - Each title should be a plain, action-oriented sentence
 - Each body should explain: what's missing, why it matters for the vision, rough approach
 - Reference blocking issues by number in depends array
+- When metrics indicate a systemic problem conflicting with VISION.md (slow CI, high blocked ratio, disk pressure), create an optimization issue even if not explicitly in VISION.md
 
 If there are no gaps, output exactly: NO_GAPS
 
diff --git a/supervisor/supervisor-poll.sh b/supervisor/supervisor-poll.sh
index aae8c3b..e6484cc 100755
--- a/supervisor/supervisor-poll.sh
+++ b/supervisor/supervisor-poll.sh
@@ -20,6 +20,22 @@ LOCKFILE="/tmp/supervisor-poll.lock"
 PROMPT_FILE="${FACTORY_ROOT}/supervisor/PROMPT.md"
 PROJECTS_DIR="${FACTORY_ROOT}/projects"
 
+METRICS_FILE="${FACTORY_ROOT}/metrics/supervisor-metrics.jsonl"
+
+emit_metric() {
+  mkdir -p "$(dirname "$METRICS_FILE")"
+  printf '%s\n' "$1" >> "$METRICS_FILE"
+}
+
+rotate_metrics() {
+  [ -f "$METRICS_FILE" ] || return 0
+  local cutoff tmpfile
+  cutoff=$(date -u -d '30 days ago' +%Y-%m-%dT%H:%M)
+  tmpfile="${METRICS_FILE}.tmp"
+  awk -F'"' -v cutoff="$cutoff" 'NF >= 4 && $2 == "ts" && $4 >= cutoff' \
+    "$METRICS_FILE" > "$tmpfile" && mv "$tmpfile" "$METRICS_FILE" || rm -f "$tmpfile"
+}
+
 # Prevent overlapping runs
 if [ -f "$LOCKFILE" ]; then
   LOCK_PID=$(cat "$LOCKFILE" 2>/dev/null)
@@ -30,6 +46,7 @@ if [ -f "$LOCKFILE" ]; then
 fi
 echo $$ > "$LOCKFILE"
 trap 'rm -f "$LOCKFILE" "$STATUSFILE"' EXIT
+rotate_metrics
 
 flog() {
   printf '[%s] %s\n' "$(date -u '+%Y-%m-%d %H:%M:%S UTC')" "$*" >> "$LOGFILE"
@@ -145,6 +162,16 @@ if [ "${DISK_PERCENT:-0}" -gt 80 ]; then
   fi
 fi
 
+# Emit infra metric
+_RAM_TOTAL_MB=$(free -m | awk '/Mem:/{print $2}')
+_RAM_USED_PCT=$(( _RAM_TOTAL_MB > 0 ? (_RAM_TOTAL_MB - AVAIL_MB) * 100 / _RAM_TOTAL_MB : 0 ))
+emit_metric "$(jq -nc \
+  --arg ts "$(date -u +%Y-%m-%dT%H:%MZ)" \
+  --argjson ram "${_RAM_USED_PCT:-0}" \
+  --argjson disk "${DISK_PERCENT:-0}" \
+  --argjson swap "${SWAP_USED_MB:-0}" \
+  '{ts:$ts,type:"infra",ram_used_pct:$ram,disk_used_pct:$disk,swap_mb:$swap}' 2>/dev/null)" 2>/dev/null || true
+
 # =============================================================================
 # P4-INFRA: HOUSEKEEPING — stale processes, log rotation (project-agnostic)
 # =============================================================================
@@ -197,6 +224,21 @@ check_project() {
   PENDING_CI=$(wpdb -c "SELECT count(*) FROM pipelines WHERE repo_id=${WOODPECKER_REPO_ID} AND status='pending' AND EXTRACT(EPOCH FROM now() - to_timestamp(created)) > 1800;" 2>/dev/null | xargs || true)
   [ "${PENDING_CI:-0}" -gt 0 ] && p2 "${proj_name}: CI: ${PENDING_CI} pipeline(s) pending >30min"
 
+  # Emit CI metric (last completed pipeline)
+  _CI_ROW=$(wpdb -A -F ',' -c "SELECT id, COALESCE(ROUND(EXTRACT(EPOCH FROM (to_timestamp(finished) - to_timestamp(started)))/60)::int, 0), status FROM pipelines WHERE repo_id=${WOODPECKER_REPO_ID} AND status IN ('success','failure','error') AND finished > 0 ORDER BY id DESC LIMIT 1;" 2>/dev/null | grep -E '^[0-9]' | head -1 || true)
+  if [ -n "$_CI_ROW" ]; then
+    _CI_ID=$(echo "$_CI_ROW" | cut -d',' -f1 | tr -d ' ')
+    _CI_DUR=$(echo "$_CI_ROW" | cut -d',' -f2 | tr -d ' ')
+    _CI_STAT=$(echo "$_CI_ROW" | cut -d',' -f3 | tr -d ' ')
+    emit_metric "$(jq -nc \
+      --arg ts "$(date -u +%Y-%m-%dT%H:%MZ)" \
+      --arg proj "$proj_name" \
+      --argjson pipeline "${_CI_ID:-0}" \
+      --argjson duration "${_CI_DUR:-0}" \
+      --arg status "${_CI_STAT:-unknown}" \
+      '{ts:$ts,type:"ci",project:$proj,pipeline:$pipeline,duration_min:$duration,status:$status}' 2>/dev/null)" 2>/dev/null || true
+  fi
+
   # Dev-agent health (only if monitoring enabled)
   if [ "${CHECK_DEV_AGENT:-true}" = "true" ]; then
     DEV_LOCK="/tmp/dev-agent.lock"
@@ -425,6 +467,18 @@ check_project() {
     unset DEPS_OF BACKLOG_NUMS NODE_COLOR SEEN_CYCLES DEP_CACHE
   fi
 
+  # Emit dev metric
+  _BACKLOG_COUNT=$(codeberg_api GET "/issues?state=open&labels=backlog&type=issues&limit=50" 2>/dev/null | jq 'length' 2>/dev/null || echo 0)
+  _BLOCKED_COUNT=$(codeberg_api GET "/issues?state=open&labels=blocked&type=issues&limit=50" 2>/dev/null | jq 'length' 2>/dev/null || echo 0)
+  _PR_COUNT=$(codeberg_api GET "/pulls?state=open&limit=50" 2>/dev/null | jq 'length' 2>/dev/null || echo 0)
+  emit_metric "$(jq -nc \
+    --arg ts "$(date -u +%Y-%m-%dT%H:%MZ)" \
+    --arg proj "$proj_name" \
+    --argjson backlog "${_BACKLOG_COUNT:-0}" \
+    --argjson blocked "${_BLOCKED_COUNT:-0}" \
+    --argjson prs "${_PR_COUNT:-0}" \
+    '{ts:$ts,type:"dev",project:$proj,issues_in_backlog:$backlog,issues_blocked:$blocked,pr_open:$prs}' 2>/dev/null)" 2>/dev/null || true
+
   # ===========================================================================
   # P4-PROJECT: Clean stale worktrees for this project
   # ===========================================================================