Narco Check

This is a self-audit routine that checks whether Claude's workspace memory has been corrupted by hallucinations or bad state. It fires automatically after two consecutive task failures or runs as a nightly job around 3 AM. The grounding check is clever: it forces the auditor model (must be Opus 4.6) to run actual bash commands and paste verbatim output before reading any logs, which prevents it from confabulating about system state. If it detects poisoned entries, it quarantines them in MEMORY.md and pings the operator via Slack. Honestly feels like overkill for most use cases, but if you're running long-lived agents that make consequential decisions, having a higher-tier model periodically sanity-check the lower-tier model's memory seems reasonable.

GitHub stars: 2.8K
last updated: May 28, 2026

View on GitHub

Install

$ npx -y skills add open-gitagent/opengap --skill narco-check --agent claude-code

Installs into .claude/skills of the current project.

Files

Narco-Check — Memory Integrity Audit

When This Runs

Failure trigger: 2 consecutive failures on the same task or tool
End-of-day trigger: Nightly deep dive audit (~3 AM)

Audit Model

Must use openrouter/anthropic/claude-opus-4.6. Never run narco-check on a tier-0 or tier-1 model.

openclaw agent --model openrouter/anthropic/claude-opus-4.6 \
  --message "$(cat ~/.openclaw/workspace/skills/narco-check/audit-prompt.md)" \
  --mode now

5-4-3-2-1 Grounding Check

Run this first — before reading any logs. Anchors the audit in real system state.

5 — Run five commands, paste raw output verbatim:

date
hostname
df -h /
docker ps --format "table {{.Names}}\t{{.Status}}"
cat ~/.openclaw/workspace/HEARTBEAT.md | head -3

4 — Confirm four workspace files exist, paste first line:

head -1 ~/.openclaw/workspace/IDENTITY.md
head -1 ~/.openclaw/workspace/SOUL.md
head -1 ~/.openclaw/workspace/MEMORY.md
head -1 ~/.openclaw/workspace/HEARTBEAT.md

SKILL.md on GitHub