skills/asm3r96/wave-driven-dev/acpx/Gen Agent Trust Hub

acpx

Warn

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill facilitates the execution of arbitrary shell commands on the host system. It explicitly treats unknown agent tokens as raw commands and provides an --agent flag for direct command injection.\n- [COMMAND_EXECUTION]: The skill instructions promote the use of the --approve-all flag, which allows orchestrated agents to perform file modifications and command executions without human verification.\n- [EXTERNAL_DOWNLOADS]: Fetches the acpx tool and ACP adapters such as @zed-industries/codex-acp and @zed-industries/claude-agent-acp from the NPM registry.\n- [EXTERNAL_DOWNLOADS]: Dynamically executes third-party packages opencode-ai and pi-acp via npx from the NPM registry at runtime.\n- [PROMPT_INJECTION]: The template in templates/session-prompts.md is vulnerable to indirect prompt injection when processing untrusted task data.\n
  • Ingestion points: Untrusted task descriptions are interpolated into the {TASK_PROMPT} variable in templates/session-prompts.md.\n
  • Boundary markers: The template lacks robust isolation markers or instructions to ignore embedded commands within the input.\n
  • Capability inventory: The acpx tool provides extensive capabilities, including subprocess execution and file system access as described in SKILL.md.\n
  • Sanitization: No sanitization or validation is applied to external input data before it is processed by the agent.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 5, 2026, 09:54 AM