skills/api7/a6/a6-persona-developer/Gen Agent Trust Hub

a6-persona-developer

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to download and install the a6 CLI tool directly from a remote repository: go install github.com/api7/a6/cmd/a6@latest. The source belongs to the skill's authoring organization.
  • [COMMAND_EXECUTION]: The skill contains numerous shell commands intended for the agent to guide the user in executing or to execute itself. These include API gateway configuration (a6 route create), local development setup (docker run, make docker-up), and connectivity testing (curl).
  • [INDIRECT_PROMPT_INJECTION]: As a developer persona, the skill is designed to ingest user-provided requirements to generate specific CLI commands and configurations, creating a vulnerability surface where malicious user input could attempt to influence command generation.
  • Ingestion points: User-provided API design specifications and routing requirements described in natural language.
  • Boundary markers: None identified in the prompt templates.
  • Capability inventory: Full suite of a6 CLI commands (CRUD operations on gateway resources), docker container management, and curl network operations as listed in the metadata.
  • Sanitization: No explicit sanitization or validation of user-provided paths or identifiers is defined within the skill instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 12:28 PM
Security Audit — agent-trust-hub — a6-persona-developer