Pick your role to see how Mimir maps to your day-to-day work.
Your standards, applied consistently — by every developer and every AI, every session. Task definition that takes minutes, not an afternoon. Feedback anchored in shared craft, not personal opinion.
With every sprint and every feature built, your team and their AI get sharper — because the playbook they share keeps improving.
Build a playbook with your architecture rules, coding conventions, and agent personas. AI reads it before every session. Every developer on the team gets the same context.
In Cursor, say:
Before starting a task, pull the relevant activity from the playbook. AI gets the steps, rules, and agent persona — no chain-prompting required.
In Cursor, say:
Reference the playbook, not your opinion. Easier to write. Harder to ignore.
In Cursor, say:
Export any workflow as markdown files directly into your AI workspace. Cursor reads them in every session — no manual copy-paste, no extra config.
In Cursor, say:
Uses export_workflow_to_local — dumps workflow activities as .md files into your target directory.
You've accumulated rules, workflows, and agent configs in
.windsurf/ or .cursor/ — they live on one machine,
in one repo. Encode them into a personal Mimir Playbook ("My Craft")
and your entire methodology travels with you: every machine, every IDE, every AI session pulls from the same source.
In Cursor, say:
AI reads your existing rules and workflow files, creates the playbook structure in Mimir, and maps each file into the right workflow → activity. Once released, every new machine gets your full craft via MCP — no copying folders around.
Plan the sprint without being the weakest person in the room. Issues that are consistent, complete, and comparable — every sprint, every team. Know the sprint is healthy or at risk before the team tells you it's too late.
With every sprint and every feature built, your team and their AI get sharper — because the playbook they share keeps improving.
Run a readiness assessment against the BPE playbook for your next milestone. Mimir checks that all prerequisite issues exist, dependencies are resolved, and the milestone is ready to start — before the first standup.
In Cursor, say:
AI reads the BPE workflow and creates GitHub issues for your milestone — each with context map, do-not-do list, and checkpoint command already written. Consistent structure every time.
In Cursor, say:
Retain organizational knowledge and capability. Know your experts and measure their contribution. Turn your playbooks into RFP answer drafts in under an hour.
With every sprint and every feature built, your team and their AI get sharper — because the playbook they share keeps improving.
Encode your team's hard-won patterns — workflows, activities, design decisions — in one place. AI reads it. New hires read it. The knowledge stops leaving when people leave.
In Cursor, say:
Release the playbook at v1.0. It becomes locked — AI can read it but cannot modify it. Changes require a Playbook Improvement Proposal (PIP) — reviewed and versioned like a code change.
In Mimir UI:
Export any workflow to JSON and hand it to another team. They import it into their Mimir instance — same structure, same activities, same guidance.
In Mimir UI:
Stop guessing what "doing it right" looks like. The playbook shows you, step by step. Ask the AI anything — it won't judge you, and it's read the same playbook your lead has.
With every sprint and every feature built, your team and their AI get sharper — because the playbook they share keeps improving.
Pull the relevant activity from the playbook. AI explains the steps, the rules, and what a good outcome looks like — before you start.
In Cursor, say:
Before you open a PR, run the checklist. The playbook defines what "done" means — AI reads it and tells you what's missing.
In Cursor, say:
Register to get your API token, connect your IDE, and run your first command in under 5 minutes.
Get started Back to home