feed-the-machine 1.7.6 → 1.7.8

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-plan-gate.sh

@@ -1,24 +1,43 @@
 #!/usr/bin/env bash
 # ftm-plan-gate.sh
-# PreToolUse hook for Edit/Write tools.
+# PreToolUse hook for Edit/Write tools AND mutating MCP calls.
 #
-# Checks if a plan has been presented this session before allowing code edits.
-# If no plan marker exists and the edit count is climbing, injects warnings
-# telling Claude to stop and present a plan first.
+# Checks if a plan has been presented this session before allowing code edits
+# or external API mutations. If no plan marker exists and the action count is
+# climbing, injects warnings telling Claude to stop and present a plan first.
+#
+# Gates: Edit, Write, and MCP tools that create/update/delete external resources
+# (Freshservice, Okta, Jira, Slack sends, Gmail sends, etc.)
 #
 # The marker file (~/.claude/ftm-state/.plan-presented) is created by Claude
 # when it presents a plan. Any non-empty content counts as "plan presented".
 # The file is cleaned up by the blackboard enforcer at session end.
 #
-# Hook: PreToolUse (matcher: Edit|Write)
+# Hook: PreToolUse (matcher: Edit|Write|mcp__*)
 
 set -euo pipefail
 
 INPUT=$(cat)
 TOOL_NAME=$(echo "$INPUT" | jq -r '.tool_name // ""')
 
-# Only gate Edit and Write tools
-if [[ "$TOOL_NAME" != "Edit" && "$TOOL_NAME" != "Write" ]]; then
+# Determine if this is a gated tool
+IS_GATED=false
+
+# Gate Edit and Write
+if [[ "$TOOL_NAME" == "Edit" || "$TOOL_NAME" == "Write" ]]; then
+  IS_GATED=true
+fi
+
+# Gate mutating MCP calls (create, update, delete, send, add, remove, apply, transition)
+if [[ "$TOOL_NAME" == mcp__* ]]; then
+  case "$TOOL_NAME" in
+    *create*|*update*|*delete*|*send*|*add*|*remove*|*apply*|*transition*|*commit*|*push*|*post_message*|*reply*|*modify*|*batch*|*convert*)
+      IS_GATED=true
+      ;;
+  esac
+fi
+
+if [[ "$IS_GATED" != "true" ]]; then
   exit 0
 fi
 

package/hooks/settings-template.json

@@ -4,7 +4,7 @@
   "hooks": {
     "PreToolUse": [
       {
-        "matcher": "Edit|Write",
+        "matcher": "Edit|Write|mcp__freshservice-mcp|mcp__mcp-atlassian|mcp__slack__slack_post|mcp__gmail__send|mcp__gmail__draft|mcp__git__git_commit|mcp__git__git_push",
         "hooks": [
           {
             "type": "command",

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "feed-the-machine",
-  "version": "1.7.6",
+  "version": "1.7.8",
   "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",