The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses local shell commands for environment validation and source control.
It executes git rev-parse and git remote show to verify the current branch and repository status.
It uses git add and git commit to track project progress.
It uses cat and shell redirection (heredoc) to format commit messages securely to prevent shell injection from task summaries.
[INDIRECT_PROMPT_INJECTION]: The skill processes task descriptions from a local file which could contain adversarial instructions.
Ingestion points: The agent reads .planning/[project-name]/tasks.md from disk to determine its next actions.
Boundary markers: There are no explicit markers or instructions to treat the file content as untrusted data or to ignore embedded instructions within the task list.
Capability inventory: The agent can write to any file in the workspace root, execute git commands, and trigger other subagents via gsdl-tracker-sync.
Sanitization: No sanitization or validation of the task file content is performed before processing.

gsdl-execute