The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses Bash to execute commands for knowledge management, including directory listing, file removal, and interacting with the qmd and agent-fs CLI tools.
[EXTERNAL_DOWNLOADS]: The skill provides instructions to install external Node.js/Bun packages: @tobilu/qmd for local semantic indexing and @desplega.ai/agent-fs for remote team storage.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by ingesting data from external and potentially shared sources.
Ingestion points: Content is retrieved via agent-fs search, qmd query, and local globbing of .md files in the learnings/ directory.
Boundary markers: No explicit markers or "ignore instructions" delimiters are specified for the ingested learning content.
Capability inventory: The skill can execute shell commands (bash), write files, and modify critical project files like CLAUDE.md during the promotion flow.
Sanitization: There is no evidence of content sanitization, filtering, or validation for learnings retrieved from remote or indexed sources.

learning