Skills

Skills are guided workflows that an AI agent follows to complete a task, encoding expertise as step-by-step instructions rather than code or APIs.

A skill follows the open Agent Skills format: a set of instructions, plus optional scripts and reference files, that an AI agent reads on demand. The same skill works across any skills-compatible agent. Develocity skills can draw on your Develocity data through the MCP Server.

How Skills Work

An AI agent uses skills through progressive disclosure:

  1. Discovery: At startup, the agent loads the name and description of each available skill.

  2. Activation: When your request matches a skill’s description, the agent loads that skill’s full instructions.

  3. Execution: The agent follows the instructions, running any bundled scripts and reading referenced files as needed.

The agent loads a skill’s full instructions only when a task calls for them, so it can keep many skills available at once. It selects the skill whose description matches your request, and some agents also let you start a skill directly.

Install Skills

Develocity distributes its skills through a marketplace served by your Develocity instance. After you add the marketplace in your AI agent, the published skills become available to it. See Getting Started for installation steps.

Available Skills

Develocity publishes the following skills:

Develocity Build Caching Report

Surveys your instance for the avoidable cache misses costing the most CI time, with a fix prompt for each.

Develocity Build Caching Optimizer

Diagnoses and fixes cache misses in a Gradle or Maven project, then opens a pull request with the changes.

Develocity Failure Report

Ranks the recurring build and test failures worth fixing across your instance, with a fix prompt for each.

Develocity Failure Handler Beta

Diagnoses one build failure from Develocity cross-build data, producing a root cause and a reproduction recipe.