The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation of a third-party Node.js package @tobilu/qmd from a public registry. This introduces a supply chain dependency on a repository not associated with the skill's primary author.
[COMMAND_EXECUTION]: The instructions describe executing shell commands where user input is directly interpolated (e.g., qmd query "<query>"). This pattern is susceptible to command injection if the agent does not properly escape the query string, allowing an attacker to execute arbitrary commands on the host system.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection (Category 8) because it retrieves and processes untrusted data from a knowledge base corpus.
Ingestion points: Document excerpts retrieved via the qmd query command output processed in SKILL.md.
Boundary markers: Absent. There are no instructions for the agent to use delimiters or ignore instructions embedded within the retrieved document content.
Capability inventory: The skill uses the Bash tool and interacts with the local file system via the qmd CLI tool.
Sanitization: Absent. The skill does not describe any validation or sanitization steps for the data retrieved from the knowledge base before it is presented to the agent.

rag-search