The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands (Bash, Glob, Grep) and Git utilities to perform its primary function.
It scans the workspace for configuration files such as docker-compose.yml, dashboard exports (Grafana/Metabase), and scheduled job definitions.
It utilizes git log and git blame to determine the modification history and authorship of identified analytics artifacts.
[DATA_EXPOSURE_RISK]: The skill is instructed to locate and review sensitive configuration files including docker-compose.yml and Slack webhook configurations.
This access is consistent with the stated purpose of auditing analytics delivery and infrastructure.
No instructions were found that exfiltrate this sensitive data to external domains; usage is restricted to local analysis and report generation.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from the workspace, such as SQL files and dashboard metadata, which represents a potential surface for indirect prompt injection.
Ingestion points: Workspace files including analytics/, reports/, and queries/ directories, and dashboard configuration files.
Boundary markers: None identified in the prompt instructions.
Capability inventory: Full file system access via Read, Glob, Grep, and shell execution via Bash.
Sanitization: No explicit sanitization or instruction to ignore embedded commands in the analyzed files is present.

lens-audit