The Agent Skills Directory

[PERSISTENCE]: The skill provides instructions to configure a recurring cron job (0 4 * * * python3 ~/.openclaw/workspace/scripts/memory-janitor.py) to perform automated memory archival.
[COMMAND_EXECUTION]: The skill requires the execution of local Python scripts (memory-janitor.py) with various flags (--stats, --dry-run) to manage system state and file organization.
[INDIRECT_PROMPT_INJECTION]: The memory management architecture is designed to ingest and process data from files such as MEMORY.md and memory/lessons/*.jsonl, which may contain external or untrusted content.
Ingestion points: MEMORY.md and memory/lessons/*.jsonl (found in SKILL.md architecture description).
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are specified for the archived data.
Capability inventory: Shell command execution via python3 and process isolation using setsid (specified in Core Principles).
Sanitization: No sanitization, escaping, or validation logic is described for the content being archived or recalled.

memory-management