a6-plugin-key-auth
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
a6CLI tool to manage local APISIX configurations, including creating consumers and routes. It also utilizescurlfor verifying authentication states against the local gateway (127.0.0.1). - [INDIRECT_PROMPT_INJECTION]: The skill processes user-defined configuration blocks (JSON/YAML) through heredocs and file parameters, which are then passed to system-level CLI tools. This creates an attack surface where maliciously crafted configuration data could influence the agent's actions.
- Ingestion points: The skill reads configuration data via
-fflags and stdin in thea6 consumer create,a6 credential create, anda6 route createcommands withinSKILL.md. - Boundary markers: Absent. The skill does not use specific delimiters or instructions to prevent the agent from interpreting instructions contained within the configuration data.
- Capability inventory: The skill can execute shell commands (
a6,curl) which modify proxy behavior and interact with network services. - Sanitization: No evidence of validation, sanitization, or escaping is present for the configuration payloads passed to the CLI tools.
Audit Metadata