The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes metadata from explainer.md files within knowledge bundles to drive its maintenance workflow. This creates a surface where malicious content in those files could influence agent behavior.
Ingestion points: explainer.md files in bundles targeted by the sweep command.
Boundary markers: There are no explicit delimiters or instructions for the agent to ignore instructions embedded in the source data.
Capability inventory: The skill executes git commands through subprocess.run and provides command templates for the agent to perform logs and diffs.
Sanitization: The tool lacks validation for the locator paths and other metadata extracted from bundle files before they are used in shell commands.
[COMMAND_EXECUTION]: The skill's operation depends on executing external commands.
Evidence: assets/garden.py uses subprocess.run to call git rev-parse, git status, and git diff. The references/refresh-playbook.md directs the agent to execute various git commands using parameters parsed from the bundle files.

knowledge-gardener