The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the alma CLI tool to interact with the file system and perform memory-related tasks. It manages data in the threads/ workspace directory and the ~/.config/alma/ directory. These operations are restricted to the vendor's established toolset and data paths.
[PROMPT_INJECTION]: The skill processes historical data, which introduces a potential surface for indirect prompt injection. Content from past interactions or group logs could contain instructions that influence the agent's behavior upon retrieval.
Ingestion points: Conversation archives in threads/, group chat logs in ~/.config/alma/groups/, and per-person profiles in ~/.config/alma/people/.
Boundary markers: The skill does not define specific delimiters or instructions to ignore potential commands within the retrieved memory content.
Capability inventory: The agent has access to Bash, Read, and Write tools to perform operations mediated by the alma CLI.
Sanitization: There is no evidence of content sanitization or validation for retrieved memory or log data prior to its inclusion in the agent's context.

memory-management