The Agent Skills Directory

[SAFE]: The skill implements a session management pattern that writes documentation to a local directory (.claude/sessions/). It uses relative paths and does not interact with sensitive system files or external services.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process and summarize session context, which may include data from external sources. While this presents a potential surface for indirect prompt injection, the skill acts as a documentation tool and does not provide an execution path for instructions contained within the session data. The risk is minimized by the structured nature of the handoff template.
[COMMAND_EXECUTION]: The instructions involve creating files and directories locally, which is a standard capability for development-focused AI agents. There are no patterns of arbitrary command execution or shell injection.

session-save