The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill acts on data retrieved from external Qlik Sense applications and knowledge bases, which constitutes an indirect prompt injection surface. This risk is effectively mitigated by pervasive governance rules requiring explicit user confirmation before the agent performs any create, update, or delete operations.
Ingestion points: Data enters the agent's context through tools like qlik_get_chart_data, qlik_search_knowledgebase_chunks, and qlik_get_dataset_sample (documented in references/qlik-mcp-tool-reference.md).
Boundary markers: While specific text delimiters for tool outputs are not mandated in the prompt instructions, the skill uses a structured discovery process (Search -> Describe -> Survey) to validate data sources.
Capability inventory: The skill has the capability to modify the user's Qlik environment using tools such as qlik_create_measure, qlik_update_glossary_term, and qlik_update_data_product.
Sanitization: The skill enforces strict data integrity rules, including mandatory verification of field values using qlik_get_field_values before applying filters, and prioritizes the use of governed Master Items over ad-hoc expressions.
[SAFE]: The skill incorporates explicit security guidelines to prevent data exfiltration during visualization. Specifically, references/visualization-guidelines.md instructs the agent to build self-contained HTML reports that exclude all external network calls, CDNs, and remote assets.

qlik-sense-app-analysis