The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell scripts (scripts/init.sh, scripts/sync-pull.sh, scripts/sync-push.sh) to manage its local storage directory and interact with CLI tools. These scripts use standard utilities like sed, awk, and grep to parse and update local markdown files.
[EXTERNAL_DOWNLOADS]: The skill interacts with GitHub and GitLab through their official CLI tools (gh and glab). This network activity is used to synchronize issue status and is consistent with the skill's primary purpose of cross-session continuity.
[PROMPT_INJECTION]: The synchronization mechanism presents a surface for indirect prompt injection.
Ingestion points: Data from remote issue titles and statuses is fetched by scripts/sync-pull.sh and intended to be processed by the agent or stored in context.md.
Boundary markers: No specific delimiters or instructions are used to distinguish external issue data from the skill's own instructions.
Capability inventory: The skill possesses capabilities for file modification and interaction with remote repository APIs via the included scripts and developer CLIs.
Sanitization: There is no evidence of sanitization or validation performed on the external issue content before it is integrated into the agent's context.

memory