agent-dbg
Pass
Audited by Gen Agent Trust Hub on Mar 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the agent-dbg CLI tool to launch, attach to, and manage Node.js processes.\n- [REMOTE_CODE_EXECUTION]: The eval and hotpatch commands allow for arbitrary code execution and runtime modification of the debugged process's logic.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from the debugged application.\n
- Ingestion points: Data enters the agent's context through agent-dbg state, agent-dbg vars, agent-dbg source, and agent-dbg console (found in references/commands.md).\n
- Boundary markers: No specific boundary markers or instructions to treat data as untrusted are present in the skill's instructions.\n
- Capability inventory: The skill can execute commands via agent-dbg launch, evaluate code via agent-dbg eval, and modify variables/files via agent-dbg set and agent-dbg hotpatch.\n
- Sanitization: No sanitization or validation of the inspected data is performed before it is presented to the agent.
Audit Metadata