snipcodeit · snipcodeit · Mar 1, 2026 · Mar 1, 2026 · Mar 1, 2026
@@ -106,6 +106,46 @@ fi
 # Prefix: default v1
 PREFIX="v1"
 ```
+
+**In extend mode, load existing metadata and ask for new milestones:**
+
+When `EXTEND_MODE=true`, skip the questions above and instead:
+
+```bash
+if [ "$EXTEND_MODE" = true ]; then
+  # Load existing project metadata — name, repo, stack, prefix are already known
+  PROJECT_NAME=$(python3 -c "import json; print(json.load(open('${REPO_ROOT}/.mgw/project.json'))['project']['name'])")
+  STACK=$(python3 -c "import json; print(json.load(open('${REPO_ROOT}/.mgw/project.json'))['project'].get('stack','unknown'))")
+  PREFIX=$(python3 -c "import json; print(json.load(open('${REPO_ROOT}/.mgw/project.json'))['project'].get('prefix','v1'))")
+  EXISTING_MILESTONE_NAMES=$(python3 -c "import json; p=json.load(open('${REPO_ROOT}/.mgw/project.json')); print(', '.join(m['name'] for m in p['milestones']))")
+
+  # Assemble project history context for the template generator
+  MILESTONE_HISTORY=$(python3 -c "
+import json
+p = json.load(open('${REPO_ROOT}/.mgw/project.json'))
+lines = []
+for m in p['milestones']:
+    lines.append(f\"Milestone: {m['name']}\")
+    for i in m.get('issues', []):
+        lines.append(f\"  - {i['title']} ({i.get('pipeline_stage','unknown')})\")
+print('\n'.join(lines))
+")
+
+  GSD_DIGEST=$(node ~/.claude/get-shit-done/bin/gsd-tools.cjs history-digest 2>/dev/null || echo "")
+
+  HISTORY_CONTEXT="Previous milestones and issues built so far:
+${MILESTONE_HISTORY}
+
+GSD build history (phases and decisions already made):
+${GSD_DIGEST:-No GSD history available.}"
+
+  # Ask only for the new work — different question for extend mode
+  # Ask: "What new milestones should we add to ${PROJECT_NAME}?"
+  # Capture as EXTENSION_DESCRIPTION
+
+  DESCRIPTION="Extension of existing project. Existing milestones: ${EXISTING_MILESTONE_NAMES}. New work: ${EXTENSION_DESCRIPTION}"
+fi
+```
 </step>
 
 <step name="generate_template">
@@ -144,6 +184,13 @@ The project details for generation:
 - **Repo:** `$REPO`
 - **Prefix:** `$PREFIX`
 
+<project_history>
+${HISTORY_CONTEXT:-No prior history available.}
+</project_history>
+
+When in extend mode (HISTORY_CONTEXT is populated): do NOT suggest features or systems that already
+appear in the project history above. Build new milestones that complement and extend what exists.
+
 After generating the JSON, extract it and write to a temp file:
 
 ```bash

diff --git a/.planning/quick/3-template-generation-should-use-project-h/3-PLAN.md b/.planning/quick/3-template-generation-should-use-project-h/3-PLAN.md
@@ -0,0 +1,259 @@
+---
+phase: quick-3
+plan: 01
+type: execute
+wave: 1
+depends_on: []
+files_modified:
+  - commands/project.md
+  - .claude/commands/mgw/project.md
+autonomous: true
+requirements:
+  - ISSUE-58
+must_haves:
+  truths:
+    - "In extend mode, the template generator receives a <project_history> block listing previous milestone names, issue titles, and pipeline stages"
+    - "In extend mode, the template generator receives a GSD history digest (if available) so it knows what was already built"
+    - "Generated milestones build on existing architecture rather than re-suggesting completed work"
+  artifacts:
+    - path: "commands/project.md"
+      provides: "Source-of-truth command — extend-mode gather_inputs step assembles HISTORY_CONTEXT"
+      contains: "project_history"
+    - path: ".claude/commands/mgw/project.md"
+      provides: "Deployed command — identical change"
+      contains: "project_history"
+  key_links:
+    - from: "gather_inputs extend-mode block"
+      to: "generate_template AI prompt"
+      via: "HISTORY_CONTEXT variable injected as <project_history> block"
+      pattern: "HISTORY_CONTEXT"
+---
+
+<objective>
+Inject project history context into the `generate_template` step when mgw:project runs in extend mode,
+so the AI template generator knows what milestones, issues, and systems already exist.
+
+Purpose: Prevents the generator from suggesting features already built in previous milestones,
+and ensures new milestones build on existing architecture decisions.
+
+Output: Updated `commands/project.md` and `.claude/commands/mgw/project.md` with history
+assembly and `<project_history>` injection added to the extend-mode section of `gather_inputs`.
+</objective>
+
+<execution_context>
+@/home/hat/.claude/get-shit-done/workflows/execute-plan.md
+</execution_context>
+
+<context>
+## Pattern to copy: mgw:issue history injection
+
+From `commands/issue.md` (spawn_analysis step), the established pattern is:
+
+```bash
+# Gather GSD project history for context (if available):
+HISTORY=$(node ~/.claude/get-shit-done/bin/gsd-tools.cjs history-digest 2>/dev/null || echo "")
+```
+
+Then inject into the AI prompt as:
+```
+<project_history>
+${HISTORY}
+</project_history>
+```
+
+## Where to add it in project.md
+
+The change goes inside the `gather_inputs` step's `EXTEND_MODE=true` block, after loading
+existing metadata (PROJECT_NAME, STACK, PREFIX, EXISTING_MILESTONE_NAMES) and before
+assembling the final DESCRIPTION variable. This new block assembles `HISTORY_CONTEXT` and
+is later injected into the `generate_template` step's AI prompt.
+
+### Current extend-mode block (lines ~130-148 in commands/project.md):
+
+```bash
+if [ "$EXTEND_MODE" = true ]; then
+  PROJECT_NAME=$(python3 ...)
+  STACK=$(python3 ...)
+  PREFIX=$(python3 ...)
+  EXISTING_MILESTONE_NAMES=$(python3 ...)
+
+  # Ask only for the new work — different question for extend mode
+  # Ask: "What new milestones should we add to ${PROJECT_NAME}?"
+  # Capture as EXTENSION_DESCRIPTION
+
+  DESCRIPTION="Extension of existing project. Existing milestones: ${EXISTING_MILESTONE_NAMES}. New work: ${EXTENSION_DESCRIPTION}"
+fi
+```
+
+### After the change, extend-mode block becomes:
+
+```bash
+if [ "$EXTEND_MODE" = true ]; then
+  PROJECT_NAME=$(python3 ...)
+  STACK=$(python3 ...)
+  PREFIX=$(python3 ...)
+  EXISTING_MILESTONE_NAMES=$(python3 ...)
+
+  # Assemble project history context for the template generator
+  MILESTONE_HISTORY=$(python3 -c "
+import json
+p = json.load(open('${REPO_ROOT}/.mgw/project.json'))
+lines = []
+for m in p['milestones']:
+    lines.append(f\"Milestone: {m['name']}\")
+    for i in m.get('issues', []):
+        lines.append(f\"  - {i['title']} ({i.get('pipeline_stage','unknown')})\")
+print('\n'.join(lines))
+")
+
+  GSD_DIGEST=$(node ~/.claude/get-shit-done/bin/gsd-tools.cjs history-digest 2>/dev/null || echo "")
+
+  HISTORY_CONTEXT="Previous milestones and issues built so far:
+${MILESTONE_HISTORY}
+
+GSD build history (phases and decisions already made):
+${GSD_DIGEST:-No GSD history available.}"
+
+  # Ask only for the new work — different question for extend mode
+  # Ask: "What new milestones should we add to ${PROJECT_NAME}?"
+  # Capture as EXTENSION_DESCRIPTION
+
+  DESCRIPTION="Extension of existing project. Existing milestones: ${EXISTING_MILESTONE_NAMES}. New work: ${EXTENSION_DESCRIPTION}"
+fi
+```
+
+### generate_template step: inject HISTORY_CONTEXT into the AI prompt
+
+In the `generate_template` step, the project details block that currently reads:
+
+```
+The project details for generation:
+- **Project name:** `$PROJECT_NAME`
+- **Description:** `$DESCRIPTION`
+- **Stack:** `$STACK`
+- **Repo:** `$REPO`
+- **Prefix:** `$PREFIX`
+```
+
+Becomes (add the history block after the project details — only rendered in extend mode):
+
+```
+The project details for generation:
+- **Project name:** `$PROJECT_NAME`
+- **Description:** `$DESCRIPTION`
+- **Stack:** `$STACK`
+- **Repo:** `$REPO`
+- **Prefix:** `$PREFIX`
+
+<project_history>
+${HISTORY_CONTEXT:-No prior history available.}
+</project_history>
+
+When in extend mode (HISTORY_CONTEXT is populated): do NOT suggest features or systems that already
+appear in the project history above. Build new milestones that complement and extend what exists.
+```
+
+`HISTORY_CONTEXT` is only set when `EXTEND_MODE=true`, so the block gracefully falls back to
+"No prior history available." in normal (non-extend) mode — no conditional needed.
+</context>
+
+<tasks>
+
+<task type="auto">
+  <name>Task 1: Add history assembly to extend-mode block and inject into generate_template prompt</name>
+  <files>
+    commands/project.md
+    .claude/commands/mgw/project.md
+  </files>
+  <action>
+Make two targeted edits to `commands/project.md` (source of truth):
+
+**Edit 1: gather_inputs step — add HISTORY_CONTEXT assembly inside the EXTEND_MODE block**
+
+Find the line:
+```
+  DESCRIPTION="Extension of existing project. Existing milestones: ${EXISTING_MILESTONE_NAMES}. New work: ${EXTENSION_DESCRIPTION}"
+```
+
+Insert the following block immediately BEFORE that line (after EXISTING_MILESTONE_NAMES is set and before DESCRIPTION is assembled):
+
+```bash
+  # Assemble project history context for the template generator
+  MILESTONE_HISTORY=$(python3 -c "
+import json
+p = json.load(open('${REPO_ROOT}/.mgw/project.json'))
+lines = []
+for m in p['milestones']:
+    lines.append(f\"Milestone: {m['name']}\")
+    for i in m.get('issues', []):
+        lines.append(f\"  - {i['title']} ({i.get('pipeline_stage','unknown')})\")
+print('\n'.join(lines))
+")
+
+  GSD_DIGEST=$(node ~/.claude/get-shit-done/bin/gsd-tools.cjs history-digest 2>/dev/null || echo "")
+
+  HISTORY_CONTEXT="Previous milestones and issues built so far:
+${MILESTONE_HISTORY}
+
+GSD build history (phases and decisions already made):
+${GSD_DIGEST:-No GSD history available.}"
+```
+
+**Edit 2: generate_template step — inject HISTORY_CONTEXT into the AI prompt**
+
+Find the block that ends with:
+```
+- **Prefix:** `$PREFIX`
+```
+(just before "After generating the JSON, extract it and write to a temp file:")
+
+Append after `- **Prefix:** \`$PREFIX\``:
+
+```
+<project_history>
+${HISTORY_CONTEXT:-No prior history available.}
+</project_history>
+
+When in extend mode (HISTORY_CONTEXT is populated): do NOT suggest features or systems that already
+appear in the project history above. Build new milestones that complement and extend what exists.
+```
+
+Then apply the IDENTICAL edits to `.claude/commands/mgw/project.md`.
+
+Note: `.claude/commands/mgw/project.md` is an older version — apply the two targeted edits
+to its equivalent locations (same step names: `gather_inputs` extend-mode block and
+`generate_template` project details block). The exact line numbers may differ from the source file.
+  </action>
+  <verify>
+    <automated>grep -n "HISTORY_CONTEXT" /hd1/repos/mgw/.worktrees/issue/58-template-generation-should-use-project-h/commands/project.md && grep -n "HISTORY_CONTEXT" /hd1/repos/mgw/.worktrees/issue/58-template-generation-should-use-project-h/.claude/commands/mgw/project.md && grep -n "project_history" /hd1/repos/mgw/.worktrees/issue/58-template-generation-should-use-project-h/commands/project.md && grep -n "project_history" /hd1/repos/mgw/.worktrees/issue/58-template-generation-should-use-project-h/.claude/commands/mgw/project.md</automated>
+  </verify>
+  <done>
+    Both files contain HISTORY_CONTEXT assembly (MILESTONE_HISTORY + GSD_DIGEST) in the extend-mode block,
+    and both contain a &lt;project_history&gt; injection in the generate_template prompt.
+  </done>
+</task>
+
+</tasks>
+
+<verification>
+After task completion:
+1. `grep -c "HISTORY_CONTEXT" commands/project.md` returns >= 3 (set, used in HISTORY_CONTEXT=, and referenced in prompt)
+2. `grep -c "HISTORY_CONTEXT" .claude/commands/mgw/project.md` returns >= 3
+3. `grep "project_history" commands/project.md` shows the XML tag in the generate_template section
+4. `grep "project_history" .claude/commands/mgw/project.md` shows the XML tag in the generate_template section
+5. `grep "MILESTONE_HISTORY" commands/project.md` confirms the python3 block that reads project.json milestones
+6. `grep "history-digest" commands/project.md` confirms the gsd-tools call
+</verification>
+
+<success_criteria>
+- extend mode assembles MILESTONE_HISTORY from project.json (milestone names + issue titles + pipeline stages)
+- extend mode calls gsd-tools history-digest for GSD artifact context
+- both are combined into HISTORY_CONTEXT variable
+- HISTORY_CONTEXT is injected as &lt;project_history&gt; block in the generate_template AI prompt
+- template generator receives explicit instruction not to re-suggest already-built features
+- changes applied identically to both commands/project.md and .claude/commands/mgw/project.md
+</success_criteria>
+
+<output>
+No SUMMARY.md required for quick plans. Changes go directly to the two command files.
+</output>
@@ -139,6 +139,26 @@ if [ "$EXTEND_MODE" = true ]; then
   PREFIX=$(python3 -c "import json; print(json.load(open('${REPO_ROOT}/.mgw/project.json'))['project'].get('prefix','v1'))")
   EXISTING_MILESTONE_NAMES=$(python3 -c "import json; p=json.load(open('${REPO_ROOT}/.mgw/project.json')); print(', '.join(m['name'] for m in p['milestones']))")
 
+  # Assemble project history context for the template generator
+  MILESTONE_HISTORY=$(python3 -c "
+import json
+p = json.load(open('${REPO_ROOT}/.mgw/project.json'))
+lines = []
+for m in p['milestones']:
+    lines.append(f\"Milestone: {m['name']}\")
+    for i in m.get('issues', []):
+        lines.append(f\"  - {i['title']} ({i.get('pipeline_stage','unknown')})\")
+print('\n'.join(lines))
+")
+
+  GSD_DIGEST=$(node ~/.claude/get-shit-done/bin/gsd-tools.cjs history-digest 2>/dev/null || echo "")
+
+  HISTORY_CONTEXT="Previous milestones and issues built so far:
+${MILESTONE_HISTORY}
+
+GSD build history (phases and decisions already made):
+${GSD_DIGEST:-No GSD history available.}"
+
   # Ask only for the new work — different question for extend mode
   # Ask: "What new milestones should we add to ${PROJECT_NAME}?"
   # Capture as EXTENSION_DESCRIPTION
@@ -184,6 +204,13 @@ The project details for generation:
 - **Repo:** `$REPO`
 - **Prefix:** `$PREFIX`
 
+<project_history>
+${HISTORY_CONTEXT:-No prior history available.}
+</project_history>
+
+When in extend mode (HISTORY_CONTEXT is populated): do NOT suggest features or systems that already
+appear in the project history above. Build new milestones that complement and extend what exists.
+
 After generating the JSON, extract it and write to a temp file:
 
 ```bash