adr-create
Pass
Audited by Gen Agent Trust Hub on Jul 30, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes local shell and filesystem tools (Bash, Write, Edit, Glob) to manage project documentation. These actions are aligned with the skill's stated purpose of ADR management.
- [INDIRECT_PROMPT_INJECTION]: The skill includes a data ingestion surface where results from mcp__plugin_ruflo-core_ruflo__memory_search are incorporated into new ADR files. While no boundary markers or sanitization logic are defined to isolate this external content, no malicious behavior is directed.
- [SAFE]: All database operations and memory search tools (mcp__plugin_ruflo-core_ruflo__agentdb_* and mcp__plugin_ruflo-core_ruflo__memory_*) are used within the context of the agent's environment for structured record keeping.
Audit Metadata