The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute a bash command to log metadata after completing a workflow. This command uses the echo utility to append a JSON string to ~/.claude/LIFEOS/MEMORY/SKILLS/execution.jsonl and executes a subshell for the date utility. If placeholders like 8_WORD_SUMMARY are replaced with unsanitized user input, it could lead to command injection.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest untrusted user requests and context, processing them through multiple "lenses" to generate binary-testable criteria.
Ingestion points: Problem/Request and Context inputs defined in Workflows/Explore.md, and customization files from ~/.claude/LIFEOS/USER/CUSTOMIZATIONS/SKILLS/IterativeDepth/.
Boundary markers: None specified in the instructions to separate user data from agent logic.
Capability inventory: The skill uses TaskCreate and TaskUpdate platform tools, and performs file-write operations via the bash logging command.
Sanitization: No explicit sanitization or validation of external content is mentioned before interpolation into prompts or shell commands.
[PERSISTENCE]: The skill implements a persistence mechanism by maintaining a local execution log in the user's home directory, allowing it to track history across different sessions.

IterativeDepth