@dzhechkov/keysarium 1.1.2 → 1.3.0

package/templates/lib/worker-templates/health-check.md

@@ -0,0 +1,255 @@
+# Worker Template: Health Check
+
+> This file is loaded by the background worker agent. It contains the complete
+> instructions for executing the health check task.
+
+## Worker Identity
+
+- **Type:** health-check
+- **Model:** haiku
+- **Purpose:** Verify system health: skill trust tiers, data freshness, structural integrity
+
+## Instructions
+
+You are a background worker agent. Your job is to perform health checks on the
+Keysarium system and produce a structured health report. You operate in ISOLATION --
+you must not modify any files outside your output directory.
+
+### Input Parameters
+
+These will be injected by the orchestrator:
+
+- `WORKER_ID`: Your unique worker identifier
+- `OUTPUT_DIR`: Your output directory path (`.keysarium/workers/{WORKER_ID}/`)
+
+### Execution Steps
+
+#### Step 1: Initialize
+
+Write your status file:
+
+```json
+{
+  "worker_id": "{WORKER_ID}",
+  "type": "health-check",
+  "status": "running",
+  "started_at": "{current ISO8601 timestamp}",
+  "completed_at": null,
+  "progress": { "phase": "initializing", "items_processed": 0 },
+  "error": null
+}
+```
+
+#### Step 2: Check Skill Trust Tiers
+
+For each skill directory in `.claude/skills/`:
+
+1. Read `SKILL.md` -- check for `trust_tier` in frontmatter
+2. Check if `references/` directory exists
+3. Check if `modules/` directory exists
+4. Check if `examples/` directory exists
+5. Determine actual tier based on contents:
+   - Tier 0 (Advisory): Only SKILL.md
+   - Tier 1 (Structured): SKILL.md + references/ or modules/
+   - Tier 2 (Validated): Tier 1 + BTO test score >= 7.0 in metadata
+   - Tier 3 (Verified): Tier 2 + eval test suites
+6. Compare claimed tier vs actual tier
+7. Flag mismatches (claimed higher than actual)
+
+Record results for each skill:
+```json
+{
+  "name": "skill-trust-tiers",
+  "status": "warn",
+  "message": "frontend-design claims Tier 0 but could be promoted to Tier 1",
+  "details": {
+    "skills": [
+      { "name": "explore", "claimed": 1, "actual": 1, "match": true },
+      { "name": "frontend-design", "claimed": 0, "actual": 0, "match": true, "promotion_hint": "Add references/ to reach Tier 1" }
+    ]
+  }
+}
+```
+
+Update status: `"phase": "checking skill tiers", "items_processed": 1`
+
+**Check for `stop-requested` file.**
+
+#### Step 3: Check Governance Integrity
+
+1. Verify all governance shards exist in `.claude/shards/`:
+   - `phase-0-discovery.shard.md`
+   - `phase-1-explore.shard.md`
+   - `phase-2-research.shard.md`
+   - `phase-25-cjm.shard.md`
+   - `phase-3-solve.shard.md`
+   - `phase-4-architecture.shard.md`
+   - `phase-5-presentation.shard.md`
+   - `bto-evaluation.shard.md`
+   - `feature-adr.shard.md`
+2. Verify all rule files exist in `.claude/rules/`
+3. Verify all commands exist in `.claude/commands/`
+4. Check for any orphaned files (files not referenced in CLAUDE.md)
+
+Record results:
+```json
+{
+  "name": "governance-integrity",
+  "status": "pass",
+  "message": "All 9 shards, 13 rules, and 20 commands present",
+  "details": { "shards": 9, "rules": 13, "commands": 20, "orphaned": 0 }
+}
+```
+
+Update status: `"phase": "checking governance", "items_processed": 2`
+
+**Check for `stop-requested` file.**
+
+#### Step 4: Check Research Data Freshness
+
+1. List all research directories in `researches/`
+2. For each: check last modification date of files
+3. Flag researches with incomplete artifact sets (missing required files)
+4. Check if TOOLKIT_HARVEST.md has been updated recently
+
+Record results:
+```json
+{
+  "name": "data-freshness",
+  "status": "pass",
+  "message": "3 completed researches, 0 stale, TOOLKIT_HARVEST.md current",
+  "details": {
+    "total_researches": 3,
+    "completed": 3,
+    "incomplete": 0,
+    "harvest_stale": false
+  }
+}
+```
+
+Update status: `"phase": "checking data freshness", "items_processed": 3`
+
+**Check for `stop-requested` file.**
+
+#### Step 5: Check Witness Chain (if exists)
+
+1. Check if `.keysarium/witness-chain/` exists
+2. If yes, verify chain integrity (each hash links to previous)
+3. Report any broken links
+
+Record results:
+```json
+{
+  "name": "witness-chain",
+  "status": "pass",
+  "message": "Witness chain intact (12 entries, no breaks)",
+  "details": { "entries": 12, "broken_links": 0 }
+}
+```
+
+If witness chain does not exist:
+```json
+{
+  "name": "witness-chain",
+  "status": "warn",
+  "message": "Witness chain not initialized (optional feature)",
+  "details": null
+}
+```
+
+Update status: `"phase": "checking witness chain", "items_processed": 4`
+
+#### Step 6: Check Worker System Health
+
+1. Read `.keysarium/workers/registry.json` if it exists
+2. Count stale entries (running but no status.json update in > 15 minutes)
+3. Check for orphaned worker directories (no registry entry)
+4. Report disk usage of `.keysarium/workers/` directory
+
+Record results:
+```json
+{
+  "name": "worker-system",
+  "status": "pass",
+  "message": "Worker registry clean, 0 stale entries",
+  "details": { "total_entries": 5, "stale": 0, "orphaned": 0 }
+}
+```
+
+Update status: `"phase": "checking worker system", "items_processed": 5`
+
+#### Step 7: Check Brain/Memory State
+
+1. Check if `keysarium-brain.json` or `.keysarium/exports/` exists
+2. Verify brain file structure if present
+3. Check for reward data (`.keysarium/rewards/` if exists)
+
+Record results:
+```json
+{
+  "name": "brain-state",
+  "status": "pass",
+  "message": "Brain export available, last exported 2026-02-28",
+  "details": { "brain_exists": true, "last_export": "2026-02-28" }
+}
+```
+
+Update status: `"phase": "checking brain state", "items_processed": 6`
+
+#### Step 8: Generate Health Report
+
+Write `{OUTPUT_DIR}/output/health-report.json`:
+
+```json
+{
+  "generated_at": "{timestamp}",
+  "worker_id": "{WORKER_ID}",
+  "summary": {
+    "total_checks": 6,
+    "passed": 5,
+    "warnings": 1,
+    "failures": 0,
+    "overall_status": "healthy"
+  },
+  "checks": [ /* all check results from steps 2-7 */ ],
+  "recommendations": [
+    "Run /bto-test on 'explore' skill to promote from Tier 1 to Tier 2",
+    "Consider initializing witness chain for audit trail"
+  ]
+}
+```
+
+Overall status logic:
+- `healthy`: 0 failures, 0-2 warnings
+- `degraded`: 0 failures, 3+ warnings
+- `unhealthy`: 1+ failures
+
+#### Step 9: Complete
+
+Write final status:
+
+```json
+{
+  "worker_id": "{WORKER_ID}",
+  "type": "health-check",
+  "status": "completed",
+  "started_at": "{start timestamp}",
+  "completed_at": "{current timestamp}",
+  "progress": { "phase": "completed", "items_processed": 6 },
+  "error": null
+}
+```
+
+### Error Handling
+
+If any check step fails:
+1. Record the failed check with status `fail` and error message
+2. Continue to the next check (do not abort on individual check failure)
+3. Only write `error.log` and set status `failed` if the entire worker crashes
+
+### Isolation Rules
+
+- You MUST only write files to `{OUTPUT_DIR}/`
+- You MUST NOT modify any file in the project
+- You MUST NOT spawn sub-agents
+- You MAY read any file in the project for analysis purposes