data-science plugin
Work-specific capabilities for the Black Duck Data Science team: LLM Gateway management, Atlassian/Confluence/JIRA workflows, data analytics patterns, and content creation guidelines.
Note: This plugin contains internal tooling references, org-specific IDs, and Black Duck infrastructure details. Not intended for use outside the Black Duck / Synopsys context.
Skills
professional-context
Team and org context: Andrew Bolster (Sr Manager R&D Engineering), Tim Cave, Conor Norris, reports to Drew Streib. Primary projects (LLM Gateway, SuperDuck, Databricks, Service-MCP, MCP Integrations). MCP tool configuration — Atlassian cloudId/space IDs, Service-MCP, WorkIQ. Tool preferences and key constraints.
/data-science:professional-context
gateway-management
LLM Gateway CLI operations via scripts/manage.py. User listing, targeted lookup, user creation (with automatic bd-staff team assignment on core), budget bumps, usage summary. Budget request workflow via WorkIQ Teams queries. Alternative data source via Service-MCP SQL. Budget warning thresholds.
/data-science:gateway-management
atlassian-workflows
Confluence and JIRA operations. Page creation with mandatory ai-generated label. Weeknote publication (personal space ~bolster, parent folder 841482874). URL verification via CQL before linking. JIRA issue creation/search/transition. Formatting standards. Large-query agent delegation guidance.
/data-science:atlassian-workflows
data-analytics
Black Duck security data product analytics. Three-catalog architecture (data_product_marts, data_product_intermediate, data_product_staging). Scale context (~4K customers → ~2.8M findings → ~20M scans). Common gotchas (duplicate records, VARIANT columns requiring STRING cast). Analysis workflow, Spark SQL standards, common query patterns including LLM Gateway.
/data-science:data-analytics
content-creation
Weeknote and blog post creation. Two-stage workflow (research in ~/WEEKNOTE.md, editorial for publishing). Weeknote structure (Successes/Failures/Blockers/In-Flight). Formatting requirements, date calculation, LLM Gateway metrics sourcing. Monthly blog post structure, YAML frontmatter. WorkIQ query patterns for research. Mermaid Gantt chart pitfalls.
/data-science:content-creation
project-management
JIRA project management for the DS project (cloudId: 739838e2-f328-4f14-a533-3f7d49323638). Ticket reference format (DS-###), when to create tickets, JQL search patterns, issue transitions. Blocker and in-flight formatting conventions. WorkIQ for Teams communication lookup.
/data-science:project-management
databricks-query
Execute SQL against the BD Databricks workspace and retrieve actual results. SQL Statement Execution API 2.0 pattern (not jobs/notebooks), workspace config (profile blackduck_general, warehouse ID), synchronous vs async query modes, wait_timeout constraints, result format, auth recovery.
/data-science:databricks-query
polaris
Add Black Duck Polaris SAST+SCA security scanning to any repo. One-time setup commands (gh variable set / gh secret set), generic polaris.yml template, GitHub Actions workflow with SARIF upload, local scripts/polaris-scan.sh with auto Bridge CLI download, .gitignore additions, and language-specific notes (standard Python, UV inline scripts workaround, non-Python projects, Makefile false-positive).
/data-science:polaris
Validation
cd plugins/data-science
claude plugin validate .
