Lythe

Slack, Teams & tickets

Run Fraser from Slack or Microsoft Teams and turn findings into engineering work in Jira or Linear.

Fraser is a product teammate that lives where your team already works: Slack or Microsoft Teams. You brief it, it runs the simulation, and it reports back in the thread, then turns what it found into engineering work in Jira or Linear.

Adding the teammate#

Install Fraser into your Slack or Microsoft Teams workspace and invite it to a channel your team already watches, an #product-qa or #eng-quality channel works well. Anyone in that channel can start a run; results post back where the whole team can see them.

You control what the teammate can reach. Connecting Slack or Teams does not grant Fraser access to your codebase or deploy pipeline. See Authorization & limits.

Starting a run#

Message Fraser and it runs a simulation, there's no run plan to review and no approval step, it just runs:

@Fraser test the signup flow at app.example.com,
goal: create a workspace on the team plan and invite a teammate.

When the run finishes, the findings come back in the thread.

Connecting Jira and Linear#

Authorize Fraser against the Jira or Linear workspace you want tickets to land in, and choose the default project or board. Each connection is scoped to the project you pick, and you can disconnect it at any time.

Fraser writes only what you allow: it creates and updates the tickets it files. It does not touch your repositories, branches, or deploys.

Turning a finding into engineering work#

From a findings thread, send any finding to your backlog:

@Fraser file finding 2 to Linear (Product board)

The ticket carries:

  • Where: the screen or step the finding names.
  • What happened: the observed behavior.
  • Suggested fix: the concrete change to consider.
  • A link back to the run it came from.

A person still reviews the ticket and ships the change. Fraser files and suggests; it does not merge or deploy on its own.

A workable loop#

  1. A build goes out.
  2. Someone messages Fraser in Slack or Teams to re-run the core flows.
  3. Findings post back in the thread within the run.
  4. The ones that matter become tickets in Jira or Linear, suggested fix attached.
  5. Your team picks them up in the tool they already use.

Related#

Was this page helpful?