openclaw-debugger

Warn

Audited by Gen Agent Trust Hub on May 14, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes several shell commands to inspect system state, including openclaw agent --status, lsof, tail, cat, ls, and curl.
  • [DATA_EXFILTRATION]: The skill accesses sensitive data sources. It instructs the agent to read and display a portion of the ANTHROPIC_API_KEY environment variable and read ~/.openclaw/providers.json, which likely contains API credentials or endpoints.
  • [PROMPT_INJECTION]: The skill has an Indirect Prompt Injection surface because it ingests untrusted data from log files and user-defined configurations.
  • Ingestion points: Gateway and agent logs (~/.openclaw/logs/gateway.log, ~/.openclaw/agents/<name>/logs/agent.log) and the agent's SOUL.md configuration file.
  • Boundary markers: Absent. The skill does not provide delimiters or instructions to ignore commands found within the processed files.
  • Capability inventory: The agent can execute shell commands (tail, cat, mv, curl) and inspect system state.
  • Sanitization: Absent. There is no evidence of filtering or validation for the content read from the file system.
Audit Metadata
Risk Level
MEDIUM
Analyzed
May 14, 2026, 08:22 AM
Security Audit — agent-trust-hub — openclaw-debugger