feat: add /nemp:cortex and /nemp:decay commands

Add Cortex intelligence layer with full report, status, --apply, --fuse,
--chains, --history, insight, resolve, learn, simulate, trust, validate,
and goal subcommands. Add decay engine command for vitality recalculation
and auto-archiving. Update save, recall, context, foresight, health
commands with cortex field support and backward compatibility.

Author: SukinShetty

CLAUDE.md

@@ -1,6 +1,6 @@
 ## Project Context (via Nemp Memory)
 
-> Auto-generated by Nemp Memory. Last updated: 2026-02-17 00:00
+> Auto-generated by Nemp Memory. Last updated: 2026-02-26 16:20
 
 ### Project Info
 
@@ -16,21 +16,32 @@
 |-----|-------|
 | **stack** | Claude Code Plugin — Markdown commands, JSON storage, agent tracking, access logs, token compression |
 | **storage-format** | JSON files — .nemp/memories.json (project), ~/.nemp/memories.json (global). Plus access.log audit trail and MEMORY.md index. |
-| **project-structure** | commands/ (16 .md files), skills/ (nemp-memory/SKILL.md), .claude-plugin/ (plugin.json, marketplace.json, hooks/), .nemp/ (config, memories, access.log), assets/ (logo/, images/, demos/) |
+| **project-structure** | commands/ (20 .md files incl. cortex.md+decay.md), skills/, .claude-plugin/, .nemp/ (memories, cortex, episodes, archive, evolution.log), assets/ |
 
 ### Features
 
 | Key | Value |
 |-----|-------|
-| **commands-available** | activity, auto-capture, auto-sync, context, export, forget, init, list, list-global, log, recall, recall-global, save, save-global, suggest, sync |
+| **commands-available** | activity, auto-capture, auto-sync, context, cortex, decay, export, forget, foresight, health, init, list, list-global, log, recall, recall-global, save, save-global, suggest, sync |
 
 ### Other
 
 | Key | Value |
 |-----|-------|
 | **author** | Sukin Shetty (contact@nemp.dev, github.com/SukinShetty) |
-| **version** | 0.2.0 |
+| **version** | 0.3.0 |
 | **license** | MIT |
 | **test-compression** | NextAuth.js auth with Google + GitHub OAuth providers |
 
+### Cortex Status
+
+| Key | Value |
+|-----|-------|
+| **cortex-version** | Tier 1-4 complete — tracking, classification, intelligence, cognitive |
+| **cortex-files** | cortex.json (chains/goals/conflicts), episodes.json, archive.json, evolution.log |
+| **cortex-commands** | /nemp:cortex, /nemp:cortex status, insight, resolve, learn, simulate, trust, validate, goal, --apply, --fuse, --chains, --history |
+| **cortex-hooks** | save (type inference + contradiction check), recall/context/foresight (vitality tracking + episodes) |
+| **trust-score** | Not yet calculated (run /nemp:cortex status) |
+| **last-reflection** | None yet — runs after 5 sessions |
+
 ---

commands/context.md

@@ -118,6 +118,54 @@ For each (key, value) in memories:
 2. **Key contains** - Key contains search term or expanded keyword
 3. **Value contains** - Value contains search term or expanded keyword
 
+### 4b. Update Vitality for All Returned Memories
+
+For each memory that matches and will be displayed to the user, update its vitality counters:
+
+1. If the memory lacks cortex fields, initialize with defaults first:
+   - `type`: `"fact"`
+   - `confidence`: `{"score": 0.65, "source": "agent-inferred", "reason": "Pre-cortex memory"}`
+   - `vitality`: all counters set to 0, `score`: 50, `state`: "active", `trend`: "stable", `last_read`: null, `decay_rate`: 0.01
+   - `links`: `{"goals": [], "conflicts": [], "supersedes": null, "superseded_by": null, "causal": []}`
+2. For each matched memory, update:
+   ```
+   vitality.reads += 1
+   vitality.reads_7d += 1
+   vitality.reads_30d += 1
+   vitality.last_read = <current ISO-8601 timestamp>
+   ```
+3. Recalculate `vitality.score` using the formula:
+   ```
+   vitality = (
+     (reads_7d × 15) +
+     (reads_30d × 3) +
+     (foresight_load_ratio × 20) +
+     (agent_reference_ratio × 25) +
+     (update_frequency × 10) +
+     (goal_link_active × 15) -
+     (correction_events × 10) -
+     (days_since_last_read × decay_rate)
+   )
+   clamped to 0-100
+   ```
+   Where:
+   - `foresight_load_ratio` = foresight_loads / (foresight_loads + foresight_skips), default 0 if both are 0
+   - `agent_reference_ratio` = agent_references / reads, default 0 if reads is 0
+   - `update_frequency` = update_count / max(1, days_since_created)
+   - `goal_link_active` = 1 if links.goals has any active goal, else 0
+4. Set `vitality.state` based on score:
+   - 80-100: `"thriving"`
+   - 50-79: `"active"`
+   - 20-49: `"fading"`
+   - 1-19: `"dormant"`
+   - 0: `"extinct"`
+5. Write ALL updated memories back to memories.json in **one write operation** (do not write per-memory).
+
+**Log the operation:**
+```bash
+echo "[$(date -u +%Y-%m-%dT%H:%M:%SZ)] CONTEXT_READ agent=${CLAUDE_AGENT_NAME:-main} query=<query> matched=<n>" >> .nemp/access.log
+```
+
 ### 5. Display Results
 
 **When matches found:**