Over the past few months I’ve been writing about how I use AI for product work. The first post covered the philosophy: context files, opinionated prompts, and how to compose the right inputs for each task. The second added slash commands and daily summaries. The third was a hands-on setup guide. And the fourth introduced project brains for keeping complex initiatives organized.

This post covers a different kind of change. The earlier additions were incremental: more commands, better context, smoother workflows. What changed recently feels more like a threshold. The system went from a tool I invoke for specific tasks to something closer to a collaborator I dispatch to do real work. Three capabilities drove that shift: multi-agent orchestration, cross-session memory, and the encoding of domain expertise into the system itself.

Multi-Agent Workflows

The clearest example is customer escalation investigations. As a PM for data products, I regularly investigate customer-reported issues: logging gaps, data discrepancies, behavior that doesn’t match expectations. These investigations require pulling information from multiple sources and cross-referencing it all into an analysis that engineering can act on.

I built a slash command that handles this as a multi-phase workflow. When I run it with a ticket ID, here’s what happens:

1. The system reads the customer ticket, extracts the core problem, identifies which product area is involved, and classifies the issue type.
2. Three specialist agents launch simultaneously, each focused on a different data source. One searches the codebase for the relevant logic and recent changes. Another searches for related tickets and prior incidents across projects. A third checks documentation and internal wiki pages for relevant operational context.
3. A fourth agent receives the combined findings and produces database queries that can confirm or refute the working hypothesis.
4. The system combines everything into a structured analysis: issue classification, root cause anchored in code where possible, customer impact, and recommended next steps.
5. A blind validator independently re-fetches every source cited in the draft to verify the claims hold up. Then an adversarial challenger looks for alternative explanations and tests whether the classification is correct.

The output is a document I can review with an engineering colleague or paste into a chat thread. It includes a confidence assessment and a data collection status table showing what was checked and what was unavailable, along with how the analysis compensated for gaps.

The command file that orchestrates all of this isn’t prompting in the traditional sense. It defines which agents to dispatch, what information each one needs, when to wait for results before proceeding, and how to handle failures gracefully. Writing this felt more like designing a workflow than writing a prompt.

I’ve applied the same pattern to other tasks. A “fix feasibility” command evaluates whether a ticket describes a code change simple enough for a PM to implement with AI coding assistance, and produces an implementation brief if the answer is yes. The specific use cases differ, but the architecture is the same: break the problem into specialist tasks that run in parallel, then synthesize and validate the results.

Cross-Session Memory

AI conversations are stateless by default. Every new session starts from zero, which means re-explaining context that should already be established. Over a few weeks of working on the same projects, this friction adds up.

I addressed this with a four-layer memory system:

• The first layer is stable facts: a compact file that captures the current state of all active work, including project status, recent decisions, and environment constraints. This is the primary orientation file. When I start a session, the AI reads it and immediately knows what’s in flight.
• The second is a session log: a reverse-chronological list of handoff notes. Each entry records what happened in a session and what threads remain open. The last three entries give enough context to pick up where I left off.
• Third, a corrections file. This holds behavioral fixes for things the AI consistently gets wrong. It’s a staging area that should shrink over time as fixes get promoted elsewhere.
• And finally, a decisions log: a cross-cutting record of decisions that don’t belong to a specific project. Each entry captures context and rationale so I don’t relitigate settled questions.

Two commands manage this. /session-start loads all four files and presents a brief summary of current state and recent sessions. /session-end reviews the conversation, writes a handoff note, and then checks whether any learnings should be promoted to infrastructure.

“Promote to infrastructure” means taking something learned during a session and baking it into the files the agent actually reads. A correction about how to handle a specific edge case in escalation investigations might start in the corrections file, then get promoted into the escalation command or a domain skill once it’s validated. The corrections file shrinks over time as that knowledge moves into the right places.

This creates a loop where the system improves its own instructions. I approve every change, so it’s not self-modifying in a creepy way. But in practice each work session can make the next one slightly better, and the compound effect over weeks is noticeable.

Domain Expertise

The earlier posts described skills like pm-thinking, which applies product methodology (problem-first thinking, measurable outcomes) to any PM-related conversation. That’s useful, but generic. It works the same way regardless of what product you’re building.

The bigger shift was building skills that encode institutional knowledge about specific products. I now have skills for each major product area my team owns: log delivery, analytics, audit logs, alerting, and data pipelines. Each skill contains the product’s architecture and common failure modes, along with which code repositories to search and which database tables hold relevant data.

This is what makes the multi-agent workflows useful. When the code investigator agent examines an escalation about missing logs, the domain skill tells it which service handles job state and which repository contains the delivery pipeline. It also flags recent architectural changes that might be relevant. Without that context, the agent produces plausible-sounding analysis that misses the specific details engineering needs.

Now every investigation that uses a skill validates or extends the knowledge it contains, and /session-end catches insights that should be added back.

How The Work Changes

The biggest change is in my own role. It’s gone from “write the right prompt” to “design the right process.” The escalation command is a workflow with phases, dependencies, and validation steps, and thinking about it that way beats trying to pack everything into a single conversation. A few other things I’ve noticed:

• Validation has to be built in. The blind validator exists because agents make mistakes. They cite files that don’t exist, mischaracterize what code does, or draw conclusions the evidence doesn’t support. Catching those issues before they reach anyone else is the whole point.
• Cross-session memory requires discipline. The system only works if I run /session-end after substantive sessions and keep stable facts current. When I skip it, the next session starts cold and I lose the compounding benefit. Automation helps, but the commitment to maintain the memory is mine.
• And domain skills need regular maintenance. Products change. Code gets refactored, pipelines get rearchitected. Skills that aren’t periodically updated drift from reality. I haven’t solved this well yet. It’s still a manual process of noticing when a skill’s knowledge is stale and updating it.

The system still makes mistakes. Multi-agent workflows are more thorough than single-prompt conversations, but they’re not infallible. The confidence assessment in the escalation output exists because sometimes the answer is “medium confidence, we couldn’t confirm this from the available data.” That honesty about limitations is more useful than false certainty.

Where This Is Going

I’m sure the specific commands and skills will look different in six months as I learn what works and what doesn’t. But the underlying pattern feels durable: compose specialist agents with deep domain context, validate their output, and feed learnings back into the system.

I’ve published updated files to the Product AI Public repo, including the session memory commands and a generalized version of the multi-agent escalation workflow. If you’re building something similar, those might be useful starting points.

None of these pieces does much on its own. It’s the way they feed each other that turned a pile of separate prompts into something I lean on every day.

I’ve written before about how I use AI for product work and how that workflow evolved with slash commands and skills. This post focuses on how to maintain context for complex, long-running projects.

The Problem: Context Fragmentation

When I’m working on a major initiative, relevant information ends up scattered everywhere: PRDs in one tool, tickets in another, meeting notes in a third, plus emails and chat threads. Every time I return to a project after a few days, I spend time reconstructing where things stand.

AI assistants can make this worse because each conversation starts fresh. I can reference files, but the model doesn’t know which files matter for this project, what decisions we’ve already made, or what questions remain open. I end up re-explaining context that should be obvious.

Project brains solve this by creating a dedicated folder for each major initiative with a standard structure that both humans and AI can navigate.

What a Project Brain Looks Like

The structure looks like this:

projects/[project-name]/
├── CONTEXT.md        # The hub: status, stakeholders, decisions, open questions
├── artifacts/        # PRDs, specs, designs, one-pagers
├── decisions/        # Decision logs with rationale and alternatives
├── research/         # Customer feedback, data analysis, technical investigation
└── meetings/         # Meeting notes related to this project

The CONTEXT.md file is a living document that answers the questions I’d need to answer every time I pick up a project:

• What’s the current status?
• Who are the stakeholders and what do they care about?
• What decisions have we made and why?
• What questions are still open?
• Where are the relevant artifacts?

When I start a conversation about a project, I point the AI to the project folder. It reads CONTEXT.md first, then can drill into specific artifacts as needed. The model immediately knows the project state without me explaining it.

A Real Example

Say I’m working on adding observability to an internal platform—something that needs coordination across multiple teams over several months. The CONTEXT.md includes:

• Quick reference table: Status, PM, engineering lead, target dates, links to the PRD and relevant tickets. Everything I’d need to orient myself.
• Problem statement: A clear articulation of the user pain. In this case: “Platform incidents go undetected until users report them, and debugging takes hours due to lack of visibility.”
• Success metrics with baselines and targets: Things like uptime targets, reduction in mean time to resolution, and alert accuracy. These anchor every conversation about scope.
• Key decisions made: A table showing what was decided, when, why, and what alternatives we considered. When someone asks “why aren’t we including component X in v1?”, the answer is already documented.
• Open questions: A checklist of unresolved issues. This prevents the AI from assuming things are settled when they’re not.
• Links: Direct paths to the PRD, spec, analysis docs, and related pages.

The decisions/ folder contains detailed decision logs for significant choices. The research/ folder holds whatever analysis informed the project direction. The meetings/ folder captures sync notes that would otherwise disappear into Gemini notes in a Google Drive… somewhere.

When to Create a Project Brain

Not every task needs this treatment. I create a project brain when:

• The work spans multiple weeks or months. Short-term tasks don’t need the overhead.
• Multiple stakeholders are involved. If I need to coordinate with other teams, having a single source of context helps.
• Decisions require documented rationale. If someone might ask “why did you do it this way?” later, a decision log is worth the investment.
• The project crosses team boundaries. Cross-functional initiatives benefit from dedicated context that doesn’t live in any one team’s space.

For simpler work, I use a flatter folder structure with documents organized by type. Project brains are for the complex initiatives where losing the thread between sessions costs me real time.

How AI Uses Project Brains

This earns its keep when I’m working with AI on project-specific tasks. A few examples:

• Preparing for a meeting: “Read the CONTEXT.md in the [project] folder. I have a spec review meeting tomorrow. What are the open questions I should raise?”
• Drafting an update: “Based on the project context, draft a status update for leadership. Focus on progress since the start of the month and remaining blockers.”
• Decision analysis: “We need to decide whether to include [component] in scope. Read the research folder and the current CONTEXT.md. What would you recommend and why?”

By the time I’m working in it, the AI already knows the project’s history and the people involved, so its recommendations fit this specific situation instead of falling back on generic best practices.

Maintaining the Project Brain

The value depends on keeping CONTEXT.md current. I’ve found a few practices help:

• Update after significant events. When a decision is made, a meeting happens, or the status changes, update the file immediately. “I’ll do it later” means it won’t happen. LLMs are great at making these updates, so you can simply say “update relevant files based on the session we just concluded.”
• Move open questions to resolved. When a question gets answered, don’t delete it. Mark it resolved and note the answer. This preserves the reasoning trail.
• Link, don’t duplicate. CONTEXT.md should point to artifacts, not contain them. Keep PRDs in the artifacts folder. Keep meeting notes in the meetings folder. The context file is a hub, not a repository.

Scaffolding New Projects

I have a slash command that scaffolds new project brains:

/new-project platform-observability

This creates the folder structure, generates a CONTEXT.md from a template, and fills out a rough draft based on whatever context I provide. Removing the friction of setup means I’m more likely to actually use the system. You can view the command here.

The template includes the standard sections (Quick Reference, Problem Statement, Success Metrics, etc.) with placeholder text. I fill in what I know and mark other sections as TBD. Even an incomplete project brain is more useful than scattered notes.

What Surprised Me

A well-organized project brain with sparse content beats a folder full of undifferentiated documents every time, because the AI (and future me) can work with structure far more easily than with a pile of files. The decision logs have paid off the most: when someone asks why we didn’t do something, I point to the log instead of reconstructing my reasoning from memory. And while I built this for the AI, I reference these files constantly myself. Staying on top of the context keeps me oriented too, not just the assistant. The structure stays flexible too. Some projects grow extra subfolders like research/customer-interviews/, others need fewer.

This approach requires discipline to maintain, and the upfront setup takes time. But for complex initiatives where context fragmentation is a real problem, project brains have been worth the investment. The AI becomes a more useful collaborator when it has access to the same context I do.

I’m still iterating on the structure. I suspect the template will look different six months from now as I learn what sections actually get used and which ones I skip every time. I’m not trying to get the folder structure perfect. I just want to stop losing context between conversations, so each time I come back to a project I can build on what I already know.

A quick post on LinkedIn about interviewing a candidate who used real-time AI got more engagement than is usual for me. And as often happens when something goes semi-viral, some folks took issue with what I said, so I want to expand on the point I was trying to make (it wasn’t that “AI is cheating”).

Here’s what I wrote:

I had my first experience interviewing a candidate who used real-time AI today. If you’re someone who uses AI daily, it’s so easy to spot. The pause before the answer, the constant eyes flicking to the other screen, the perfectly-manicured 3-point answer…

Friends, just don’t do this. It’s too easy to spot, and it will also set you up for failure, because it might get you a job that you’re not a good fit for, which is bad for everyone.

Use AI in your job, for sure. But don’t use it to get the job. The interview process is about you. Be you.

One response called this “absolutely myopic” (I had to double check I didn’t accidentally post on Hacker News) and asked why candidates shouldn’t use AI if it allows for “a better, more creative answer.” Another suggested that if candidates will use AI on the job anyway, then the “real you” isn’t going to be working, so what’s the difference?

Let’s dig into this.

What interviews are actually for

I don’t interview people to test whether they can produce a good answer to a question. I interview people to understand how they think, what they’ve actually done, and whether we’ll work well together.

When I ask “Tell me about a time you had to make a difficult prioritization decision,” I’m not looking for the theoretically optimal framework. I want to hear your story. The messy details and the trade-offs you wrestled with. The thing you got wrong and what you learned from it. AI can’t give me that. It can only give me a polished summary of what prioritization frameworks exist.

One commenter put it well: “It’s about both the company and the individual, so you will often talk about their real experience, what they did, how they felt, what did they learn, digging deeper into their real experience to find out the interesting things that could make them a good match.”

AI might help you phrase things more clearly. But if it’s generating your answers, you’re hiding the very thing I’m trying to evaluate.

The fit problem

Here’s the part that didn’t seem to land: using AI to get a job you’re not qualified for is bad for you.

Let’s say the AI-assisted interview works. You get hired. Now what? You show up on day one, and the expectations are set based on how you performed in those interviews. But that wasn’t you. That was a performance enhanced by a tool you won’t have in the same way during actual work conversations, whiteboard sessions, and quick chat exchanges where people expect you to just… know things.

I’ve seen what happens when there’s a mismatch between interview performance and actual capability. It’s not a fun experience for anyone, least of all the person who’s now struggling in a role they weren’t ready for. One person called it “artificial buzzword ventriloquism” in the comments. Harsh, but not wrong.

It’s about context, not absolutes

A few commenters suggested that interviews should evolve to assume AI assistance, since that’s how people will actually work. One person wrote: “By prohibiting AI during interviews, the interview environment diverges from actual job conditions and fails to evaluate a critical skill: the ability to effectively use one of the most powerful productivity tools available today.”

I think there’s something to this. In fact, our interview process includes a take-home assessment where we explicitly encourage candidates to use AI. We want to see how they approach a problem, how they structure their thinking, and yes, how they use modern tools to get to a good answer. That’s a legitimate skill worth evaluating.

But that’s different from what happened in my interview, where someone was clearly trying to hide their AI usage while answering questions about their past experience. That’s not “using AI as a tool.” That’s using AI as a mask.

---

I think candidates should absolutely use AI to prepare for interviews: research the company, practice answering common questions, refine their resume.

But in the interview itself, when I’m asking about your experience and your thinking, I need to hear from you. Not because AI is cheating, but because the whole point is to figure out if you are the right fit for this role and this team. If I can’t evaluate that, we can’t make a good hiring decision. And that’s bad for both of us.

This is the hands-on companion to How I Use AI for Product Work and How My AI Product Second Brain Evolved. Those posts explain the philosophy; this one gets you to a working OpenCode setup in 30 minutes.

By the end, you’ll have a folder structure, three working slash commands, and a context file that makes the AI useful.

Prerequisites

You’ll need OpenCode installed and configured. Follow the installation guide to get started. If you can run opencode in your terminal and get a response, you’re ready.

Step 1: Create the Folder Structure

Create a new directory for your product second brain. I recommend keeping it in a git repo so you can version your prompts over time.

mkdir -p product-ai/{context,prompts/pm,.opencode/command}
cd product-ai
git init

Your structure should look like this:

product-ai/
├── .opencode/
│   └── command/       # Slash commands live here
├── context/           # Personal context files
├── prompts/
│   └── pm/            # PM-specific prompts
├── AGENTS.md          # Instructions for OpenCode
└── opencode.jsonc     # OpenCode config

Step 2: Create Your Context File

The context file tells the AI who you are and how you work. Create context/about-me.md:

# About Me

## Role
[Your title] at [Company], working on [your area].

## What I Care About
- [Your product philosophy, e.g., "Start with the problem, not the solution"]
- [Your working style, e.g., "Bias toward shipping and learning"]
- [Your communication preferences, e.g., "Direct feedback, no hedging"]

## Current Focus
- [Project or initiative 1]
- [Project or initiative 2]

Keep this updated as your focus changes. The more specific you are, the more useful the AI becomes.

Step 3: Create AGENTS.md

This file tells OpenCode how to behave in your repo. Create AGENTS.md in the root:

# Product AI Second Brain

Read this file before responding.

## Who I Am
Read `context/about-me.md` for personal context.

## Slash Commands
Run these by typing the command in OpenCode:

| Command | Purpose |
|---------|---------|
| `/prd` | Review a PRD |
| `/debate` | Stress-test a product idea |
| `/okr` | Review OKRs |

Step 4: Add Your First Commands

Slash commands are markdown files in .opencode/command/. Each command file is a thin wrapper that points to a full prompt file in prompts/pm/. This separation keeps commands simple while allowing prompts to be detailed and shareable.

Command 1: /debate (Stress-test an Idea)

Create .opencode/command/debate.md:

---
description: Stress-test a product idea with pro vs skeptic debate
---

# Product Debate

Read these files before proceeding:
- `prompts/pm/debate-product-idea.md` - **REQUIRED: Full debate framework**
- `context/about-me.md` - Your product beliefs and context

## Your Task

$ARGUMENTS

## Instructions

Then create the prompt file prompts/pm/debate-product-idea.md with your full debate methodology. The basic structure: define two personas (a Visionary who argues for the idea and a Skeptic who pokes holes), have them debate, then synthesize the strongest arguments from both sides.

Command 2: /okr (Review OKRs)

Create .opencode/command/okr.md:

---
description: Review OKRs for clarity and outcome-orientation
---

# OKR Review

Read these files before proceeding:
- `prompts/pm/review-okrs.md` - **REQUIRED: Full OKR review framework**
- `context/about-me.md` - Your product beliefs and context

## Your Task

$ARGUMENTS

## Instructions

Then create prompts/pm/review-okrs.md with your OKR criteria. Mine checks for outcome-orientation (are these outputs or outcomes?), measurability, and whether the key results actually ladder up to the objective.

Command 3: /prd (Review a PRD)

Create .opencode/command/prd.md:

---
description: Review a PRD for completeness and clarity
---

# PRD Review

Read these files before proceeding:
- `prompts/pm/review-prd.md` - **REQUIRED: Full PRD review framework**
- `context/about-me.md` - Your product beliefs and context

## Your Task

$ARGUMENTS

## Instructions

Then create prompts/pm/review-prd.md with your PRD review criteria.

Step 5: Try It Out

Navigate to your product-ai directory and run OpenCode:

cd product-ai
opencode

Test each command:

/debate Should we build a self-serve dashboard for customers to debug their own issues?

/okr [paste your OKRs here]

What’s Next

Once you’re comfortable with this setup, consider adding:

• More commands: /retro for retrospectives, /feedback for drafting colleague feedback
• Skills: Methodology files that OpenCode loads automatically based on context (see the skills documentation)
• Daily summaries: A /today command that summarizes what you worked on
• Project-specific context: Folders for major initiatives with their own context files

Start small and add complexity as you identify repeated workflows. If you find yourself doing the same thing more than twice, it’s probably worth automating.

If you build something useful on top of this, I’d love to hear about it!