DE//AGENT
Explore

Roadmap · stage 13/13

Agentic Data Engineering

Stage 13 of 13 · 26 resources · 1 categories

What this stage is

Agentic Data Engineering is stage 13 of 13 in the DE//AGENT roadmap. It currently holds 26 resources: 10 tutorials, 9 references, 4 mcp servers across 1 category.

What it teaches

Skills, MCP, agents and hooks as first-class data engineering tools.

What you can do after this stage

  • Install and use a skill
  • Wire an MCP server least-privilege
  • Build a review agent with guardrails

Exit gate

Gate: You can delegate real work to an agent with bounded blast radius. The stage deliverable: agent-sql-reviewer.

Gates are self-assessed — the roadmap does not track your progress; it tells you what to be able to do before moving on.

Categories in this stage

Registry resources for this stage

Sorted by composite score — provenance, completeness and trust, per the scoring model:

ResourceTypeDifficultyVerificationScore
Filesystem MCP ServerMCP serverbeginnercommunity47.5
Git MCP ServerMCP serverbeginnercommunity47.5
Fetch MCP ServerMCP serverbeginnercommunity46
GitHub MCP ServerMCP serverbeginnercommunity41.1
MCP Tool Contract AuditorSkilladvancedcommunity40.5
Agent-in-the-Loop ReviewPatternintermediatecommunity35.75
Keep an agent read-only until a human approvesTutorialbeginnercommunity34.15
What makes an agent — model, tools, loop, guardrailsTutorialbeginnercommunity34.15
Wire the core MCP servers for a data repoTutorialbeginnercommunity34.15
Audit an MCP server's tool contractsTutorialintermediatecommunity34.15
Hooks — deterministic gates around an agentTutorialintermediatecommunity34.15
Inside a skill — SKILL.md, references, examples, the loading protocolTutorialintermediatecommunity34.15
Onboard an agent with the DE Data ToolkitTutorialintermediatecommunity34.15
Capstone — the guarded agent checklistTutorialadvancedcommunity34.15
Production gotchas — untrusted content, destructive actions, permission scopeTutorialadvancedcommunity34.15
When an agent does the wrong thing — contain, audit, postmortemTutorialadvancedcommunity34.15
DE Data ToolkitPluginbeginnercommunity34.1
Agent Skills (open format)Reference—community27.7
MCP Registry (official)Reference—community27.7
Model Context Protocol SpecificationReference—community27.7
Daily Life as a Jenkins AdministratorReference—community23.55
GSoC 2026 Final Blog: How It Was to Be a Contributor at JenkinsReference—community23.55
Jenkins and AI Agents: The Trusted Engine for Faster Software DeliveryReference—community23.55
kitfunso/hippo-memoryReference—community23.55
Monitoring MWAA-orchestrated ETL pipelines with Amazon OpenSearch ServiceReference—community23.55
oaslananka/kicad-mcp-proReference—community23.55

Documented learning objectives

  • Read a tool schema and predict its side effects
  • Classify tools by blast radius
  • Write a least-privilege MCP configuration
  • Separate agent read/recommend from write/execute
  • Encode the safety policy as hooks
  • Keep a human gate on anything irreversible
  • Name the four parts of an agent — model, tools, loop, guardrails
  • Identify each part in a real registry agent

Continue the progression

← Full roadmap · All learning topics