Skip to content

skilldrop·docs·How-to

Use skilldrop with Jira

Feed Jira tickets and epics into skilldrop skills — bug triage, story splitting, implementation loops, and release notes.

skilldrop has no Jira API integration — it works with text you paste or have your agent retrieve via an MCP server. The patterns below show exactly what to copy and what to say.

Bug triage from a Jira ticket

Copy the ticket's summary, description, steps to reproduce, and environment fields. Then:

Here is a Jira bug ticket:

Summary: [paste]
Description: [paste]
Steps to reproduce: [paste]
Environment: [paste]

Run bug-triage.

bug-triage returns a severity assessment, a root-cause hypothesis, and a recommended next step. Paste that back into the Jira ticket as a comment or into a linked Confluence page.

Story splitting from an epic

Copy the epic's goal and any child stories already defined. Then:

Here is a Jira epic:

Goal: [paste epic summary and description]
Existing stories: [paste titles, or "none yet"]

Run user-story-splitter.

The skill returns independently shippable slices with acceptance criteria for each. Create one Jira sub-task per slice, using the acceptance criteria as the Definition of Done.

Implementation from a story

Take one Jira story and its acceptance criteria. Reference the ticket ID so the session output is traceable:

Jira ticket: PROJ-123
Title: [paste story title]
Acceptance criteria:
- [paste each criterion]

Run feature-implement-loop.

feature-implement-loop treats the acceptance criteria as its gate — each checkpoint confirms criteria are still being met before the next iteration starts. When done, link the resulting PR back to PROJ-123.

Release notes from a sprint

At the end of a sprint, export the completed issues list (Jira's sprint report has a "Copy" option, or use jira sprint list --completed). Then:

Here are the completed Jira tickets from sprint 42:

PROJ-101: [title]
PROJ-108: [title]
PROJ-115: [title]

Run release-notes.

release-notes groups changes by type and writes a changelog entry. Copy the result into your release ticket or Confluence release page.

Using the Jira MCP

If the Jira MCP is installed in Claude Code, you can ask the agent to fetch ticket text directly instead of copying:

Fetch PROJ-123 from Jira and run bug-triage on it.

The agent retrieves the ticket and passes it to the skill inline.

See also: Use skills with MCP servers

Edit this page on GitHub · View source