The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands, including cp for file manipulation, git for version control, and the letta CLI for agent management.
[DATA_EXFILTRATION]: The migration process involves exporting and copying agent memory blocks, which typically contain sensitive historical data or user information. The use of git push results in the network transmission of this data to a configured remote repository.
[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by facilitating the ingestion of memory content from an external source (another agent) without validation.
Ingestion points: Memory files are imported from /tmp/letta-memory-<agent-id> and ~/.letta/agents/ as described in the migration steps of SKILL.md.
Boundary markers: The instructions do not define delimiters or provide specific prompts to the agent to disregard instructions embedded within the migrated data.
Capability inventory: The agent has permissions to perform filesystem writes, execute management CLI commands, and perform network synchronization via Git.
Sanitization: The skill lacks any steps for sanitizing, filtering, or verifying the integrity of the content being moved between agents.

migrating-memory