Phase 02 · Define
Spec → epic with code context
How it works
A spec becomes a build-ready epic – anchored in the real code, so engineering accepts it without ping-pong.
- Reads the spec from Confluence and the code context from the repo
- Generates acceptance criteria from real code paths and data shapes
- Creates the epic in Jira with affected components and dependencies
- Links stories and flags open quality and edge-case questions
Roles
Head of product & engineering
Teklens apps & integrations
JiraConfluenceGitHubSlack
Example prompts & actions
Copy one of the following prompts and paste it into the Teklens chat. Adapt the details to your use case.
Turn the spec at [link] into an epic with acceptance criteria and code context.
Which components does this spec touch, and how risky is the change?
Generate the stories for epic [epic name], including edge cases.
Related use cases
Three use cases closely connected to this one.
Prioritisation by impact and feasibility.Prioritises features by business impact × technical feasibility – grounded in the real code instead of gut feeling.View use case A roadmap anchored in Jira and code.Connects strategic priorities with Jira, dependencies and real engineering reality – and flags it when the two drift apart.View use case PRD expertWrites code-grounded PRDs that reference real modules, data and constraints.View use case
From the PM Lab
Articles that explore this use case in depth.
- Skill files: the next AI skill after promptingRead post
- The anatomy of a Jira ticket for AI product management in agentic engineeringRead post
- Discovery stays human, execution becomes the spec: how product managers work with AI in brownfieldRead post
- The MVP for AI: test the wizard before you build the machineRead post
Give Teklens a real job.
Connect Jira and your code – and let Teklens work on something real.