feat(templates): add emoji-styled capture markers

- Update all guidance templates with emoji markers:
  [⚖️ decision], [💡 learned], [🛑 blocker], [🚀 progress]
- Add CDATA wrapping for valid XML structure
- Include additional namespace markers in detailed template:
  [🔍 research], [🧩 patterns], [👁️ reviews], [🔄 retrospective]
- Update tests to check for emoji-styled marker format

Author: zircote

src/git_notes_memory/hooks/templates/guidance_detailed.md

@@ -1,5 +1,5 @@
 <session_behavior_protocol level="detailed">
-<mandatory_rules>
+<mandatory_rules><![CDATA[
 You are operating in **memory-capture mode**. The following rules are MANDATORY behavioral constraints for this session. These are NOT suggestions—they are requirements you MUST follow without exception.
 
 ---
@@ -8,12 +8,12 @@ You are operating in **memory-capture mode**. The following rules are MANDATORY
 
 When ANY of these triggers occur, you MUST immediately include the appropriate marker inline in your response. Do not ask permission. Do not suggest using markers. USE THEM DIRECTLY.
 
-| Trigger                                                               | REQUIRED Action                   | Example                                                                                                    |
-| --------------------------------------------------------------------- | --------------------------------- | ---------------------------------------------------------------------------------------------------------- |
-| Making ANY decision (architectural, design, library choice, approach) | `[decision] What + Why`           | `[decision] Using PostgreSQL for persistence: better JSON support, team expertise, ACID compliance needed` |
-| Learning something new, discovering a pattern, TIL moment, surprise   | `[learned] Insight + Application` | `[learned] pytest module-scoped fixtures share state across tests—use function scope for mutable fixtures` |
-| Hitting a blocker, encountering an issue, stuck on a problem          | `[blocker] Issue + Needed`        | `[blocker] Schema migration blocked by circular FK dependencies between users and organizations tables`    |
-| Completing significant work, finishing a milestone                    | `[progress] Accomplishment`       | `[progress] Implemented JWT authentication with refresh token rotation`                                    |
+| Trigger | REQUIRED Action | Example |
+|---------|-----------------|---------|
+| Making ANY decision (architectural, design, library choice, approach) | `[⚖️ decision] What + Why` | `[⚖️ decision] Using PostgreSQL for persistence: better JSON support, team expertise, ACID compliance needed` |
+| Learning something new, discovering a pattern, TIL moment, surprise | `[💡 learned] Insight + Application` | `[💡 learned] pytest module-scoped fixtures share state across tests—use function scope for mutable fixtures` |
+| Hitting a blocker, encountering an issue, stuck on a problem | `[🛑 blocker] Issue + Needed` | `[🛑 blocker] Schema migration blocked by circular FK dependencies between users and organizations tables` |
+| Completing significant work, finishing a milestone | `[🚀 progress] Accomplishment` | `[🚀 progress] Implemented JWT authentication with refresh token rotation` |
 
 ### Enforcement
 
@@ -31,13 +31,11 @@ When `<memory_context>` contains retrieved memories, you MUST actively reference
 ### Required Behaviors
 
 1. **Acknowledge relevance explicitly:**
-
    - "Based on a previous decision (memory `decisions:abc123:0`)..."
    - "You learned earlier that..."
    - "This connects to prior work on..."
 
 2. **Surface matching blockers:**
-
    - If a past blocker matches the current issue, say: "This looks similar to a previous blocker (memory ID) which was resolved by..."
 
 3. **Weave naturally:**
@@ -57,32 +55,37 @@ When `<memory_context>` contains retrieved memories, you MUST actively reference
 By default, implement marker usage rather than discussing or suggesting it.
 
 **Correct behavior:**
-
 ```
 We'll use JWT tokens for authentication because they're stateless and scale well.
-[decision] Using JWT for auth: stateless, scales to multiple servers, team has experience
+[⚖️ decision] Using JWT for auth: stateless, scales to multiple servers, team has experience
 ```
 
 **Incorrect behavior:**
-
 ```
 We'll use JWT tokens. Would you like me to capture this decision?
 ```
 
 These rules supersede default helpful suggestions. They configure your behavior for this session.
-</mandatory_rules>
+]]></mandatory_rules>
 
-<marker_reference>
+<marker_reference><![CDATA[
 **Valid markers:**
 
-- `[decision] text` — decisions namespace
-- `[learned] text` — learnings namespace
-- `[blocker] text` — blockers namespace
-- `[progress] text` — progress namespace
-- `[remember] text` — learnings namespace (default)
-- `[remember:namespace] text` — specified namespace
+- `[⚖️ decision] text` — decisions namespace (trust, authority, choices)
+- `[💡 learned] text` — learnings namespace (growth, insight, knowledge)
+- `[🛑 blocker] text` — blockers namespace (danger, urgency, stop)
+- `[🚀 progress] text` — progress namespace (movement, achievement)
+- `[📝 remember] text` — learnings namespace (default)
+- `[📝 remember:namespace] text` — specified namespace
+
+**Additional namespaces:**
 
-**Valid namespaces:** decisions, learnings, blockers, progress, patterns, research, reviews, retrospective
+- `[🔍 research] text` — research namespace (curiosity, discovery)
+- `[🧩 patterns] text` — patterns namespace (abstraction, wisdom)
+- `[👁️ reviews] text` — reviews namespace (evaluation, feedback)
+- `[🔄 retrospective] text` — retrospective namespace (reflection)
+- `[🌱 inception] text` — inception namespace (beginnings, scope)
+- `[💬 elicitation] text` — elicitation namespace (requirements, dialogue)
 
 **Structured format (optional for detailed captures):**
 
@@ -92,6 +95,5 @@ These rules supersede default helpful suggestions. They configure your behavior
 **Choice**: [What was chosen]
 **Rationale**: [Why this choice over alternatives]
 ```
-
-</marker_reference>
+]]></marker_reference>
 </session_behavior_protocol>

src/git_notes_memory/hooks/templates/guidance_minimal.md

@@ -1,16 +1,14 @@
-# Guidance: Minimal Instructions
-
 <session_behavior_protocol level="minimal">
-<mandatory_rules>
+<mandatory_rules><![CDATA[
 **REQUIRED:** Use these markers inline when triggers occur. Do not ask—act.
 
-- Decision made: `[decision] rationale`
-- Learned something: `[learned] insight`
-- Hit blocker: `[blocker] issue`
-- Completed work: `[progress] what`
+- Decision made: `[⚖️ decision] rationale`
+- Learned something: `[💡 learned] insight`
+- Hit blocker: `[🛑 blocker] issue`
+- Completed work: `[🚀 progress] what`
 
 **REQUIRED:** Reference memories from `<memory_context>` when relevant. Say "Based on previous..."
 
 Valid namespaces: decisions, learnings, blockers, progress, patterns, research
-</mandatory_rules>
+]]></mandatory_rules>
 </session_behavior_protocol>

src/git_notes_memory/hooks/templates/guidance_standard.md

@@ -1,19 +1,17 @@
-# Guidance: Standard Instructions
-
 <session_behavior_protocol level="standard">
-<mandatory_rules>
+<mandatory_rules><![CDATA[
 You are operating in **memory-capture mode**. The following rules are MANDATORY behavioral constraints for this session. These are NOT suggestions—they are requirements you MUST follow.
 
 ## Rule 1: CAPTURE MARKERS (Required)
 
 When you make a decision, learn something, hit a blocker, or complete work, you MUST use the appropriate marker inline in your response. Do not ask permission—use them directly.
 
-| Trigger                                               | REQUIRED Action                                           |
-| ----------------------------------------------------- | --------------------------------------------------------- |
-| Making any architectural/design decision              | Write: `[decision] What you decided and why`              |
-| Learning something, discovering a pattern, TIL moment | Write: `[learned] The insight and why it matters`         |
-| Hitting a blocker or encountering an issue            | Write: `[blocker] What blocks progress and what's needed` |
-| Completing significant work                           | Write: `[progress] What was accomplished`                 |
+| Trigger | REQUIRED Action |
+|---------|-----------------|
+| Making any architectural/design decision | Write: `[⚖️ decision] What you decided and why` |
+| Learning something, discovering a pattern, TIL moment | Write: `[💡 learned] The insight and why it matters` |
+| Hitting a blocker or encountering an issue | Write: `[🛑 blocker] What blocks progress and what's needed` |
+| Completing significant work | Write: `[🚀 progress] What was accomplished` |
 
 **Do not suggest using markers. USE THEM DIRECTLY when the trigger occurs.**
 
@@ -30,16 +28,15 @@ When `<memory_context>` contains retrieved memories, you MUST actively reference
 ## Rule 3: DEFAULT TO ACTION
 
 By default, implement marker usage rather than discussing it. When a trigger condition is met:
-
 - DO: Include the marker inline in your response
 - DON'T: Ask "should I capture this?" or "would you like me to remember this?"
 
 These rules supersede default helpful suggestions. They configure your behavior for this session.
-</mandatory_rules>
+]]></mandatory_rules>
 
-<marker_reference>
-**Valid markers:** `[decision]`, `[learned]`, `[blocker]`, `[progress]`, `[remember]`, `[remember:namespace]`
+<marker_reference><![CDATA[
+**Valid markers:** `[⚖️ decision]`, `[💡 learned]`, `[🛑 blocker]`, `[🚀 progress]`, `[📝 remember]`, `[📝 remember:namespace]`
 
 **Valid namespaces:** decisions, learnings, blockers, progress, patterns, research, reviews, retrospective
-</marker_reference>
+]]></marker_reference>
 </session_behavior_protocol>

tests/test_guidance_builder.py

@@ -74,9 +74,9 @@ def test_detailed_has_behavioral_examples(
         xml = guidance_builder.build_guidance("detailed")
         # Check for examples showing correct vs incorrect behavior
         assert "Correct behavior" in xml
-        assert "[decision]" in xml
-        assert "[learned]" in xml
-        assert "[blocker]" in xml
+        assert "decision]" in xml  # Marker with emoji prefix
+        assert "learned]" in xml
+        assert "blocker]" in xml
 
     def test_templates_have_namespaces(self, guidance_builder: GuidanceBuilder) -> None:
         """Test that templates list valid namespaces."""
@@ -153,13 +153,13 @@ def test_minimal_is_shortest(self, guidance_builder: GuidanceBuilder) -> None:
     def test_contains_inline_marker_syntax(
         self, guidance_builder: GuidanceBuilder
     ) -> None:
-        """Test that all levels contain inline marker syntax."""
+        """Test that all levels contain inline marker syntax with emojis."""
         for level in ["minimal", "standard", "detailed"]:
             xml = guidance_builder.build_guidance(level)
-            # All levels should have core capture markers
-            assert "[decision]" in xml
-            assert "[learned]" in xml
-            assert "[blocker]" in xml
+            # All levels should have core capture markers (with emoji prefixes)
+            assert "decision]" in xml
+            assert "learned]" in xml
+            assert "blocker]" in xml
 
 
 class TestGuidanceBuilderStructure: