The Agent Skills Directory

[SAFE]: The skill consists entirely of markdown documentation and LookML configuration examples. It does not include any executable scripts, binaries, or instructions for remote code execution.- [INDIRECT_PROMPT_INJECTION]: The skill provides instructions for processing user-supplied filters and parameters via Liquid templates, which creates a potential surface for indirect injection. However, the skill explicitly mitigates this risk by instructing the agent to use the | sql_quote filter for sanitization. 1. Ingestion points: User-defined filters (_filters) and parameters (_parameter_value) described in sub-skills/lookml-liquid/SKILL.md. 2. Boundary markers: The skill demonstrates the use of {% condition %} tags to delineate filter logic. 3. Capability inventory: The skill's scope is limited to defining LookML semantic models for SQL generation. 4. Sanitization: The documentation explicitly mandates using the | sql_quote Liquid filter to prevent SQL injection vulnerabilities.- [CREDENTIALS_SAFE]: No hardcoded API keys, tokens, or credentials were found. The skill provides guidance on using Looker's standard security model, including access_grant and user_attribute parameters, to manage row-level and object-level security appropriately.

lookml