oc2-bof-script-development
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill provides instructions for executing compiled binary code (Beacon Object Files) on remote implants using tasks such as
exec_bofandexec_bof_async. This is the primary intended functionality of the skill within the context of the Outflank C2 framework. - [EXTERNAL_DOWNLOADS]: The documentation references a vendor-specific JFrog repository for resolving dependencies such as
oc2-sdk-pythonvia theuvpackage manager. This is standard configuration for the vendor's software distribution. - [INDIRECT_PROMPT_INJECTION]: The skill describes an attack surface where the agent processes operator-supplied input to generate command arguments for remote execution.
- Ingestion points: Operator CLI input ingested via
CommandContext.argumentsand processed in wrapper scripts. - Boundary markers: The skill recommends using
DefaultArgumentParserto enforce structured command schemas. - Capability inventory: The skill facilitates binary execution (
schedule_task) and interaction with a C2 API (ImplantsApi,ArtifactsApi). - Sanitization: Instructions advise validating operator input and avoiding raw argument pass-through to mitigate command injection risks.
Audit Metadata