confluent-cloud-cdc-tableflow

Set up end-to-end Change Data Capture (CDC) pipelines on Confluent Cloud using Debezium source connectors, Flink for transformation, and Tableflow for data lake integration. Supports JSON_SR, Avro, and Protobuf formats. Handles schemaless topics (plain JSON without SR) and multi-event topics. This skill handles the complete workflow from database to Iceberg/Delta tables. Use this skill when users want to capture database changes and materialize them into Iceberg or Delta Lake tables via Confluent Cloud Tableflow. Trigger phrases include "CDC to Tableflow", "database to Iceberg", "database to Delta Lake", "stream database changes to data lake", "set up Tableflow pipeline", "schemaless topic to Tableflow", or "multi-event topic to Iceberg". Do NOT trigger for general CDC, Debezium, or database replication requests that do not involve Tableflow or Iceberg/Delta Lake as the destination.

Quality

88%

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Passed

No known issues

Quality

Content

77%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

This is a strong, highly actionable skill that covers a complex multi-component pipeline with clear sequencing and validation checkpoints. Its main weakness is length — several sections (CLI fallback examples, database-specific details like Oracle prerequisites, DynamoDB-specific Flink patterns) duplicate content that should live in the referenced files, inflating the token cost. The critical architecture rules section is excellent and represents exactly the kind of non-obvious operational knowledge that justifies a skill file.

Suggestions

Move the CLI fallback examples in Phase 0.1 and database-specific prerequisite details (especially Oracle XStream) to their respective reference files to reduce inline verbosity — a one-line pointer is sufficient.

Move DynamoDB-specific Flink patterns to references/flink-sql-patterns.md and reference it inline, keeping the main skill focused on the general CDC workflow.

Dimension	Reasoning	Score
Conciseness	The skill is quite long (~400+ lines) and includes some information Claude already knows (e.g., what CDC is, what Debezium does). However, much of the content is domain-specific operational knowledge (tombstone behavior, changelog mode immutability, cleanup ordering) that genuinely adds value. Some sections like the CLI fallback examples could be trimmed since Claude knows CLI patterns.	2 / 3
Actionability	The skill provides fully executable MCP tool calls, CLI commands, and Flink SQL statements with concrete parameters. Configuration examples are copy-paste ready with clear placeholder patterns. The verification table with specific tool calls and expected outputs is highly actionable.	3 / 3
Workflow Clarity	The five-phase workflow is clearly sequenced with explicit validation checkpoints at each stage (verify connector tasks, check schemas, confirm Flink table discovery, verify INSERT job status, check Tableflow status). Error recovery paths are well-defined with a troubleshooting quick-reference table and feedback loops (wait and retry patterns). The cleanup order is explicitly documented with warnings about destructive operations.	3 / 3
Progressive Disclosure	The skill references five separate reference files (database-prerequisites.md, connector-configs.md, flink-sql-patterns.md, troubleshooting.md, rest-api.md) which is good structure, but no bundle files were provided to verify they exist. The main SKILL.md itself is quite long and includes substantial inline detail (e.g., full CLI fallback examples, DynamoDB-specific patterns, Oracle prerequisites) that could be offloaded to the reference files it already points to.	2 / 3
	Total	10 / 12 Passed

Description

100%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

This is an excellent skill description that thoroughly covers specific capabilities, provides rich natural trigger terms, explicitly addresses both what and when, and includes negative trigger guidance to minimize conflicts. It uses proper third-person voice throughout and avoids vague language or buzzwords. The inclusion of supported formats, topic types, and a complete workflow description makes this highly actionable for skill selection.

Dimension	Reasoning	Score
Specificity	Lists multiple specific concrete actions: setting up CDC pipelines, using Debezium source connectors, Flink for transformation, Tableflow for data lake integration, handling schemaless topics, multi-event topics, and supporting specific formats (JSON_SR, Avro, Protobuf). Very detailed about the complete workflow from database to Iceberg/Delta tables.	3 / 3
Completeness	Clearly answers both 'what' (set up end-to-end CDC pipelines with Debezium, Flink, Tableflow, supporting multiple formats and topic types) and 'when' (explicit 'Use this skill when...' clause plus trigger phrases and a 'Do NOT trigger' exclusion clause). Exemplary completeness.	3 / 3
Trigger Term Quality	Excellent coverage of natural trigger phrases including 'CDC to Tableflow', 'database to Iceberg', 'database to Delta Lake', 'stream database changes to data lake', 'schemaless topic to Tableflow', and 'multi-event topic to Iceberg'. Also includes explicit negative triggers to avoid false matches.	3 / 3
Distinctiveness Conflict Risk	Highly distinctive with a clear niche: CDC pipelines specifically targeting Confluent Cloud Tableflow with Iceberg/Delta Lake destinations. The explicit 'Do NOT trigger' clause for general CDC/Debezium requests without Tableflow further reduces conflict risk with related skills.	3 / 3
	Total	12 / 12 Passed

Validation

90%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 10 / 11 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (521 lines); consider splitting into references/ and linking	Warning

	Total	10 / 11 Passed

Repository: confluentinc/agent-skills
Commit: 9095afc

Reviewed: 21 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.