kata-doctor
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill performs directory renames, project documentation updates, and Git commits to manage project health. These operations are restricted to the
.planningdirectory and its subdirectories.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from project planning files to automate structural changes.\n - Ingestion points: Processes
.planning/ROADMAP.mdand.planning/STATE.mdto extract phase names and milestone data.\n - Boundary markers: No specific boundary markers or instructions to ignore embedded content are used during ingestion.\n
- Capability inventory: The skill can execute shell commands (
mv,git commit) and perform file system writes through thekata-lib.cjsutility.\n - Sanitization: The skill mitigates risks by sanitizing extracted strings using
tr -cd 'a-z0-9-'before using them as directory names in shell commands, effectively preventing command injection patterns.
Audit Metadata