Moyn

A reproducible sequence of prompts for comprehensive platform analysis across 6 repositories.

Phase 1: Repository Setup

Clone those repo via ssh into folders in this folder:

Phase 2: Repository Analysis

Great, and now lets do an exploration for each of those repos.
We'll use one subagent per repo with the following goals:

Understand the repo, what it does, business logic, request flows,
api up/downstream surface/usage, architecture, datamodels,
data sources, queues, etc - and then each subagent should write
a markdown report file ANALYSIS.md into the repo

Phase 3: System Architecture

Ok, great - and now lets build a full architecture analysis,
like - how are those repos working together in production.
Write a markdown file

Phase 4: Project History & Contributors

Ok, now lets use subagents again for each repo, with the goal
to create a STORY.md in their repo, where they tell the history
of the project, what happened, what are bottlenecks / problematic
spots in the codebase. What's the story and trajectory of the
project, who are the key contributors, whos mostly doing lowkey changes.

Phase 5: Challenges & Pain Points

What are the painpoints / challenges based on our analysis so far

Then:

Write this as markdown file

Phase 6: Microservices Deep Dive

And now tell me about the microservices, how many are there exactly,
and what's their responsibility / domain / relation to others.
Are these truly encapsulating their domain, or do we have crossfire
request flows. Lets use subagents to explore this, and write an
analysis for each microservice found

Phase 7: Frontend Analysis

And how many frontends are there, what's their stack and which
services / apis are they calling? Lets do a similar analysis there

Phase 8: User Journeys

Great, now lets analyse the actual user journeys / jobs to be done.
Do it thoroughly, end2end. Do a first pass and then use subagents
to dive deep on each. Goal is to have one file per subagent as
output (markdown) and a highlevel summary

Phase 9: Admin Flows

Ok, then lets try general purpose agents to do user journeys
analyses / jobs to be done for the admin flows. First pass
highlevel and then deepdive with those subagents producing a file each

Phase 10: Liberation Strikes (Strategic Interventions)

Now lets build up context on @FRONTENDS.md @MICROSERVICES.md
@SYSTEM_ARCHITECTURE.md and @CHALLENGES.md

What could be "liberation strikes", as in - focused, potentially
radical, changes, where a "Special Forces" team could parachute
in and really move the needle

Then for more aggressive options:

More radical, assuming we're using all AI tools to the fullest
extent, there are no limits, throw away code, consolidate services,
think bigger without constraints

Then:

Write this as markdown file

Phase 11: Visualization & Sharing

Now lets make illustrations / infographics out of
/journeys/LIBERATION_STRIKES.md using nanobanana cli
to share this with my team

Expected Outputs

After running this playbook, you should have:

File	Description
`*/ANALYSIS.md`	Technical analysis for each repository
`*/STORY.md`	Git history narrative for each repository
`SYSTEM_ARCHITECTURE.md`	How repos work together in production
`CHALLENGES.md`	Pain points and recommendations
`MICROSERVICES.md`	Service analysis with domain boundaries
`FRONTENDS.md`	Frontend apps, stacks, and API integrations
`journeys/*.md`	User journey documentation (8 files)
`journeys/USER_JOURNEYS.md`	High-level summary
`journeys/admin/*.md`	Admin flow documentation (6 files)
`journeys/LIBERATION_STRIKES.md`	Strategic intervention proposals