Security audit
Security checks for vulnerabilities and agentic risk
The skill is a coherent agent-debugging guide, but it includes under-scoped guidance to expose internal reasoning and persist debugging instructions in memory.
Install only if you are comfortable reviewing and constraining its debugging advice. Do not enable hidden/internal reasoning visibility, and apply memory or tool-policy changes only with explicit user intent and narrow scope.
Skill manipulates agent memory, state, or stored context. Memory corruption can alter personality, override safety rules, or cause unpredictable behavior.
print(f"Reasoning: {status['reasoning']}")
### Clear Context
If agent is stuck:
The skill recommends enabling internal reasoning visibility without user choice, necessity limits, or policy justification. This normalizes disclosure of privileged model internals and creates a direct path to exposing sensitive hidden reasoning during routine troubleshooting.
Skill allows unbounded resource consumption (API calls, storage, compute). Without rate limits or quotas, a compromised or misbehaving agent can cause denial-of-service or cost overruns.
---
name: agent-debugger
description: Debug AI agent issues systematically. Covers tool failures, infinite loops, context overflow, rate limits, and performance bottlenecks. Use when agents misbehave, loop infinitely, fail tools, hit limits, or produce unexpected outputs. Triggers on "debug", "fix agent", "agent stuck", "agent looping", "tool failed", "rate limit".
---
# Agent Debugger
The trigger list lacks clear scope constraints and includes generic phrases like "fix agent" and "debug," which may cause unintended invocation. In context, unintended activation matters because the skill contains guidance touching permissions, memory handling, filesystem inspection, and internal reasoning visibility.
The trigger list lacks clear scope constraints and includes generic phrases like "fix agent" and "debug," which may cause unintended invocation. In context, unintended activation matters because the skill contains guidance touching permissions, memory handling, filesystem inspection, and internal reasoning visibility.
Code scans file system directories looking for sensitive files. This could be reconnaissance for credential theft.
Verify memory setup:
ls -la ~/.openclaw/workspace/
# Should show:
# MEMORY.md
# memory/
Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.
Check file permissions:
ls -la /path/to/file
chmod 600 ~/.openclaw/workspace/sensitive.json
Review tool policies:
The skill explicitly recommends enabling "thinking: verbose" and "reasoning: on" during debugging. Exposing internal reasoning is unnecessary for most operational troubleshooting and can leak hidden chain-of-thought, policy-sensitive content, or other confidential internal state to users or logs.
Debugging instructions that surface hidden reasoning can expose sensitive model content, secrets present in context, or internal safety logic to end users or logs. In a debugging skill, this is especially risky because users may follow the recommendation during incidents, when systems are already in an unstable state.
No suspicious patterns detected.