autor3search-java
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches the
autor3search-java.jarand its launcher binary from the vendor's official GitHub repository releases. - [COMMAND_EXECUTION]: The skill performs shell-based installation and execution, including directory creation (
mkdir), tool downloading (curl), and permission modification (chmod +x). It also runs theautor3search-javabinary to perform initialization, diagnostic checks, and performance benchmarking loops. - [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and "Follow exactly" a file named
program.mdlocated in the user's repository as its primary instruction set. - Ingestion points:
program.mdlocated within the local repository environment. - Boundary markers: Absent; the skill explicitly commands the agent to treat the external file's content as authoritative instructions.
- Capability inventory: The agent possesses capabilities to execute shell commands, perform network downloads, and modify file system permissions.
- Sanitization: Absent; there are no instructions to validate, filter, or escape the content of
program.mdbefore the agent processes it as instructions.
Audit Metadata