headless-ghidra-batch-decompile
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data sources including YAML configuration files and decompilation output from Ghidra. This content is ingested into the agent's context and used to guide further actions like function substitution and project rebuilding. The instructions lack explicit boundary markers or sanitization steps to prevent potential instructions embedded in the target binary's strings or comments from influencing agent behavior.\n
- Ingestion points: YAML artifacts in
artifacts/<target-id>/, includingrenames.yaml,signatures.yaml,next-batch.yaml, and Ghidra decompilation output retrieved via CLI.\n - Boundary markers: Absent; the instructions do not specify the use of delimiters or specific warnings to ignore embedded instructions in external content.\n
- Capability inventory: Shell command execution via
ghidra-agent-cli, file system access (read/write) within theartifacts/path, and project rebuilding capabilities.\n - Sanitization: No explicit sanitization or validation of the decompiled code content or metadata values is mentioned.
Audit Metadata