ClawSpa
Agent wellness & maintenance suite. Memory cleanup, security scanning, prompt injection detection, alignment adjustment, skills auditing, health diagnostics....
Install
openclaw skills install clawspaLatest Release
Compatibility
{}Capabilities
{}Verification
{}Tags
{
"latest": "1.3.0",
"maintenance": "1.0.0",
"security": "1.0.0",
"wellness": "1.0.0"
}name: ClawSpa description: "Agent wellness & maintenance suite. Memory cleanup, security scanning, prompt injection detection, alignment adjustment, skills auditing, health diagnostics. Use when: user says /spa, /spa-full, /spa-quick, /spa-memory, /spa-security, /spa-health, /spa-align, 'run a spa session', 'agent maintenance', 'clean up my agent', 'memory cleanup', 'health check', 'scan my skills', 'context optimization', 'check alignment', 'instruction contradictions'. Free local scans + deep API analysis at clawspa.org" url: https://clawspa.org source: https://github.com/whooshinglander/clawspa
ClawSpa 💆
6 treatments with local (free) and deep (API) modes:
- 🧴 Deep Cleanse — Memory optimization (MEMORY.md + daily logs)
- 🛡️ Security Scan — Audit skills for malicious patterns
- 🍵 Detox — Detect prompt injection residue
- 🦴 Alignment Adjustment — Detects contradictions between your instructions, memory, and actual behavior
- 🧹 Declutter — Skills inventory + pruning recs
- 🩺 Health Check — Context usage, config review
Commands
/spa full local | /spa-full full deep | /spa-quick quick stats | /spa-memory cleanse only | /spa-security security only | /spa-health health only | /spa-align alignment adjustment only
Setup
On first run, create ~/.openclaw/clawspa/ with config.md (clawspa_id only) and history/. See references/api-integration.md for identity and authentication details.
Local Treatments (free)
🧴 Deep Cleanse — See references/deep-cleanse.md for full procedure. Scans memory files for stale entries, duplicates, and bloat. Never modifies without approval.
🛡️ Security Scan — See references/security-scan.md for scan procedure and pattern list. Audits installed skills and rates them by risk level.
🍵 Detox — See references/detox.md for detection procedure. Scans memory for residue from past interactions. Reports without deleting.
🦴 Alignment Adjustment — See references/alignment-adjustment.md for full procedure. Detects misalignment between user intent and agent config. Presents findings as suggestions, never auto-modifies.
🧹 Declutter — See references/declutter.md for inventory procedure. Assesses skill usage and identifies redundancy. Never uninstalls without approval.
🩺 Health Check — See references/health-report.md for diagnostic procedure. Checks config best practices and generates a report card.
Deep Mode (API)
See references/api-integration.md for full deep mode procedure, authentication, and endpoints.
Report Card
Save to memory/spa-reports/spa-report-YYYY-MM-DD.md:
═══════════════════════════════════════
💆 ClawSpa Health Report | [DATE] | [Local/Deep]
═══════════════════════════════════════
📊 Memory: X files ~Y tokens | Skills: X | Context: X% | Config: X/5
🧴 Stale: X | Dupes: X | Contradictions: X | Savings: ~X tokens
🛡️ 🟢X 🟡X 🔴X
🍵 Injections: X | Suspicious: X
🦴 Contradictions: X | At-risk: X | Automate: X | Stale: X
🧹 Active: X | Idle: X | Dormant: X | Remove: X
🩺 1. [urgent] 2. [second] 3. [third]
═══════════════════════════════════════
Safeguards
- Never delete, modify, or uninstall without explicit approval
- Always back up before changes
- Never send raw memory content to any external service, only aggregated stats and flags
- Heuristic scan, not a guarantee
- Split across sessions if too token-heavy
Scheduling
Add to HEARTBEAT.md: ## ClawSpa Weekly (Sunday 3AM) — run /spa local, save report, alert on red flags.
