feed-the-machine 1.7.7 → 1.7.9

package/ftm-mind/references/decide-act-protocol.md

@@ -91,11 +91,21 @@ If the next move will reveal new information, plan to re-enter Observe after the
 
 Act is clean, decisive execution — but execution of **approved** work only.
 
-**Pre-Act checkpoint**: Before executing anything, verify:
+**HARD GATE — Pre-Act checkpoint**: Before executing ANYTHING (Bash, MCP, Write, Edit, API calls of any kind), verify ALL of these:
 
-1. If `approval_mode` is `plan_first` or `always_ask`, did the user explicitly approve the plan?
-2. If the task involves external mutations (see Approval Gates), have you presented the specific actions and received approval?
-3. If neither condition applies, proceed.
+1. **Did you present a checkbox plan?** If the task is medium+ (forced escalation signals fired), you MUST have presented a `N. [ ] action → target` plan and received explicit user approval. "I'll do X, Y, Z" in prose is NOT a plan. Listing steps without `[ ]` checkboxes is NOT a plan. If you haven't presented one, STOP and present it now.
+2. **Did the user approve it?** Look for "go", "approve", "yes", "lgtm", or similar. If the user hasn't responded to your plan yet, WAIT. Do not start executing.
+3. **Is the plan marker written?** After approval, write to `~/.claude/ftm-state/.plan-presented` before executing. This signals to hooks that planning happened.
+4. If the task involves external mutations (see Approval Gates), have you presented the specific actions and received approval?
+5. If none of the above apply (micro/small task, no forced escalation), proceed.
+
+**The rationalization trap**: You will feel the urge to skip the plan because:
+- "The user said 'do as much as you can' — that's implicit approval" → NO. That's the task description, not plan approval.
+- "I know what needs to happen, presenting a plan is just overhead" → NO. The plan is for the USER, not for you.
+- "I'll just start with one small API call to check something" → NO. One call becomes five becomes a full execution without approval.
+- "The user seems impatient" → NO. A 30-second plan saves 10 minutes of unwanted work.
+
+**This applies to ALL execution methods** — Bash commands, MCP calls, Python scripts, curl, direct API calls. The plan-gate hook catches Edit/Write/MCP, but Bash API calls bypass it. This checkpoint is the only thing that catches those. Do not skip it.
 
 ### 1. Direct action
 

package/hooks/ftm-auto-log.sh

@@ -18,22 +18,9 @@
 FTM_STATE="$HOME/.claude/ftm-state"
 CONTEXT_JSON="$FTM_STATE/blackboard/context.json"
 
-# Check if an ftm session is active by reading context.json
-FTM_ACTIVE=$(python3 -c "
-import json, sys
-try:
-    with open('$CONTEXT_JSON') as f:
-        d = json.load(f)
-    task = d.get('current_task', {})
-    status = task.get('status', '')
-    print('1' if status not in ('', 'completed', 'none') else '0')
-except Exception:
-    print('0')
-" 2>/dev/null)
-
-if [ "$FTM_ACTIVE" != "1" ]; then
-    exit 0  # Not in active ftm session, skip
-fi
+# NOTE: Session gate removed in v1.7.9. Daily logging should happen for ALL work,
+# not just formal ftm sessions. Most productive work happens outside /ftm invocations
+# and was going untracked because context.json status was "none" or "completed".
 
 # Read payload from stdin (JSON) — extract prompt for action pattern matching
 STDIN_DATA=$(cat)
@@ -83,10 +70,20 @@ if echo "$USER_MESSAGE" | grep -qE "(I|i) .* (what|should|next|now)\?"; then
     SHOULD_LOG=true
 fi
 
-# If action detected, output logging reminder for Claude to see
+# If action detected, output logging reminder with exact instructions
+TODAY=$(date +%Y-%m-%d)
+DAILY_DIR="$HOME/.claude/ftm-ops/daily"
+DAILY_FILE="$DAILY_DIR/$TODAY.md"
+
 if [ "$SHOULD_LOG" = true ]; then
     echo ""
-    echo "[Auto-log] Detected progress update. Please log this to the ftm daily log."
+    echo "[Auto-log] Detected progress update. Append to daily log NOW:"
+    echo "  File: $DAILY_FILE"
+    if [ ! -f "$DAILY_FILE" ]; then
+        echo "  File doesn't exist yet — create it with header: # Daily Log — $TODAY"
+    fi
+    echo "  Format: - HH:MM — [what was done] (ticket/PR/system if applicable)"
+    echo "  Do this silently — don't announce it to the user."
     echo ""
 fi
 

package/hooks/ftm-session-end.sh

@@ -1,52 +1,58 @@
 #!/bin/bash
 # ftm-session-end.sh
-# Hook: Deactivate ftm session tracking when conversation ends
-# Trigger: SessionEnd
+# Hook: Deactivate ftm session tracking and ensure daily log exists when conversation ends
+# Trigger: Stop
 #
-# IMPORTANT: Must be listed AFTER ftm-session-snapshot.sh in settings.json
-# (in a separate matcher entry). The snapshot hook reads context.json status
-# to gate itself — this hook marks the session completed.
+# Always runs — not gated on ftm session state. Every session that ends should:
+# 1. Mark context.json as completed (if active)
+# 2. Ensure today's daily log file exists (create if missing)
+# 3. Remind Claude to append a session summary before exiting
 
 FTM_STATE="$HOME/.claude/ftm-state"
 CONTEXT_JSON="$FTM_STATE/blackboard/context.json"
+TODAY=$(date +%Y-%m-%d)
+DAILY_DIR="$HOME/.claude/ftm-ops/daily"
+DAILY_FILE="$DAILY_DIR/$TODAY.md"
 
-# Check if an active session exists
-IS_ACTIVE=$(python3 -c "
-import json, sys
-try:
-    with open('$CONTEXT_JSON') as f:
-        d = json.load(f)
-    task = d.get('current_task', {})
-    status = task.get('status', '')
-    print('1' if status not in ('', 'completed', 'none') else '0')
-except Exception:
-    print('0')
-" 2>/dev/null)
-
-if [ "$IS_ACTIVE" != "1" ]; then
-    exit 0
-fi
-
-# Mark session as completed in context.json
+# Mark session as completed in context.json (if it has an active task)
 python3 -c "
-import json, sys
+import json, sys, os
 from datetime import datetime
 
+ctx_path = '$CONTEXT_JSON'
+if not os.path.exists(ctx_path):
+    sys.exit(0)
+
 try:
-    with open('$CONTEXT_JSON') as f:
+    with open(ctx_path) as f:
         d = json.load(f)
 
     if 'current_task' in d and isinstance(d['current_task'], dict):
-        d['current_task']['status'] = 'completed'
+        status = d['current_task'].get('status', '')
+        if status not in ('', 'completed', 'none'):
+            d['current_task']['status'] = 'completed'
 
     if 'session_metadata' in d and isinstance(d['session_metadata'], dict):
         d['session_metadata']['last_updated'] = datetime.utcnow().strftime('%Y-%m-%dT%H:%M:%SZ')
 
-    with open('$CONTEXT_JSON', 'w') as f:
+    with open(ctx_path, 'w') as f:
         json.dump(d, f, indent=2)
-except Exception as e:
-    sys.stderr.write(f'ftm-session-end: failed to update context.json: {e}\n')
-    sys.exit(1)
+except Exception:
+    pass
 " 2>/dev/null
 
-echo "ftm session tracking deactivated (session ended)"
+# Ensure daily log directory and file exist
+mkdir -p "$DAILY_DIR"
+if [ ! -f "$DAILY_FILE" ]; then
+    echo "# Daily Log — $TODAY" > "$DAILY_FILE"
+    echo "" >> "$DAILY_FILE"
+fi
+
+# Remind Claude to write session summary to daily log
+echo ""
+echo "[Session ending] Before you finish, append a summary of this session to the daily log:"
+echo "  File: $DAILY_FILE"
+echo "  Format: ## Session — HH:MM"
+echo "  Then bullet points of what was accomplished, decisions made, and next steps."
+echo "  Do this silently — just write the file."
+echo ""

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "feed-the-machine",
-  "version": "1.7.7",
+  "version": "1.7.9",
   "description": "A brain upgrade for Claude Code — 26 skills that teach it how to think before acting, remember across conversations, debug like a war room, run plans on autopilot with agent teams, and get second opinions from GPT & Gemini. Plus 15 hooks that automate the boring stuff.",
   "license": "MIT",
   "author": "kkudumu",