AI-PHISR

Your product management brain that never forgets. Built on Claude Code + Obsidian.

5 AI Agents 21 Skills 8 Playbooks Persistent Context Zero Code
What is it
Two tools. One operating system.
AI-PHISR combines a knowledge vault with an AI engine. Your product context lives in Obsidian. Claude Code reads it, reasons with it, and writes back when a command creates an artifact or updates memory. Every interaction makes the system smarter.
📚
Obsidian Vault
Your product knowledge
in plain Markdown
+
⌨
Claude Code
AI that reads, reasons
and writes back when appropriate
=
⚙
AI-PHISR
PM operating system
with persistent memory
How it works
The flow: request to vault write-back
Every interaction follows the same loop. Context in, intelligence out, and artifacts or memory updates saved when the command is designed to do that.
1
💬
Your Request
Ask anything PM-related
➔
2
🔎
Context Load
Reads your vault automatically
➔
3
🤖
Agent Routes
Picks the right agent + skill
➔
4
⚡
Skill Executes
Generates structured output
➔
5
💾
Vault Write-back
Saves artifacts or memory updates
AI Agents
5 specialized agents, each with a job
Agents don't just respond -- they route your request to the right skill and validate the output.
🔍

Problem Framer

Turns vague requests into clear, testable problem statements. Forces clarity before solutions. Routes to Hypothesis Builder and Assumption Extractor.

⚖

Decision Architect

Structures complex decisions with criteria, trade-offs, and hidden constraints. Routes to Tradeoff Mapper and Priority Stack.

🎤

Stakeholder Translator

Translates between technical and business audiences. Builds alignment narratives. Routes to 1:1 Prep, Narrative Refiner, PM Voice Creator.

🛡

Structural Integrity Auditor

Auto-runs on every artifact. Catches logical gaps, missing assumptions, unclear success criteria, and dependency blind spots.

📖

Learner

Scans resources and surfaces relevant frameworks, case studies, and insights connected to your current product context.

Product Lifecycle
Skills mapped to your Agile loop
21 skills cover every phase of product development. Here's where each one fits.
⚙
AI-PHISR
Operating System
💡
Discovery
Hypothesis Builder
Assumption Extractor
Research Synthesizer
🎯
Strategy
Priority Stack
OKR Partner
Business Case Builder
📋
Planning
PRD Builder
Edge Case Finder
Cross-Team Mapper
🛠
Building
Build Companion
Build Review
Fire Responder
🚀
Launch
Performance Tracker
Pitch Deck Builder
Narrative Refiner
📈
Review
Sprint Summarizer
Retro Facilitator
Tradeoff Mapper
👥
People
1:1 Prep
Persona Builder
PM Voice Creator
🔄
Iterate
Continuous context
Memory persistence
Sprint state tracking
Skills
21 slash commands at your fingertips
Type a command in Claude Code. It reads your vault, runs the skill, and then either saves an artifact, updates memory, or returns a display-only output depending on the command.
/hypothesis-builderHypothesis Builder
/assumption-extractorAssumption Extractor
/priority-stackPriority Stack
/1on1-prep1:1 Prep
/prd-builderPRD Builder
/sprint-summarizerSprint Summarizer
/business-case-builderBusiness Case Builder
/retro-facilitatorRetro Facilitator
/research-synthesizerResearch Synthesizer
/persona-builderPersona Builder
/tradeoff-mapperTradeoff Mapper
/okr-partnerOKR Partner
/narrative-refinerNarrative Refiner
/cross-team-mapperCross-Team Mapper
/build-companionBuild Companion
/build-reviewBuild Review
/fire-responderFire Responder
/performance-trackerPerformance Tracker
/pitch-deck-builderPitch Deck Builder
/edge-case-finderEdge Case Finder
/pm-voice-creatorPM Voice Creator
Playbooks
8 ready-made workflows for PM moments
Playbooks chain multiple skills together. One command triggers an entire multi-step workflow, though not every playbook auto-saves a file by default.
Playbook 01

💡 New Idea

Stakeholder request comes in. Frames problem, extracts assumptions, builds testable hypothesis.

Playbook 02

🏃 Sprint Kickoff

Starts sprint with full context. Priorities aligned to bets, dependencies flagged, goals set.

Playbook 03

🚀 Launch

Go/no-go checklist. Validates hypotheses, checks dependencies, builds comms plan.

Playbook 04

🔥 Incident

Something breaks. Triage, communicate to stakeholders, resolve, schedule retro.

Playbook 05

🤝 Manager 1:1

Decision-focused agenda. No status updates. Open decisions, blockers, escalations.

Playbook 06

📅 Weekly Review

Close the week. OKR check, decisions reviewed, risks surfaced, next week planned.

Playbook 07

🎤 Pitch

Build persuasive pitch. Audience analysis, narrative arc, data-backed arguments, counter-arguments.

Playbook 08

📈 Post-Launch Data

Metrics vs. hypotheses. User feedback synthesized. Iteration recommendations generated.

The key difference
Context that compounds over time
Most AI tools forget everything between sessions. AI-PHISR remembers through its vault and memory files. Saved decisions, hypotheses, sprint updates, and related artifacts make the next interaction smarter.
❌

Regular AI Chat

Session 1: "Here's my product context..."
Session 2: "Here's my product context..."
Session 3: "Here's my product context..."
Session 4: "Here's my product context..."
Same starting point. Every. Time.
✅

AI-PHISR

Session 1: Knows your strategy + bets
Session 2: + last sprint decisions
Session 3: + validated hypotheses
Session 4: + stakeholder feedback
Compounding intelligence. Always.
Your vault
What lives inside Obsidian
Plain Markdown files organized by purpose. No databases, no APIs, no lock-in.
🎯

Strategy

Vision, Strategic Bets, OKRs

🚀

Initiatives

Active projects with status tracking

🧪

Hypotheses

Testable bets with validation status

⚖

Decisions

Decision log with context + rationale

🔍

Research

Personas, insights, user feedback

👥

Team

Stakeholders, dependencies, rituals

🏃

Sprints

Sprint goals, summaries, state

🧠

Memory

Session logs, sprint state, context index

Get started
5 days to your own PM operating system
The Half-Sprint Guide. One focused action per day. By Friday, you're running AI-PHISR on your real product.

Setup & Strategy

Fill out Setup_Worksheet.md with your product details. Run the bootstrap command.

/populate-strategy

Map Your Initiatives

Create a file for each active initiative. Add status, strategic bet, OKR link, priority.

/prd-builder

Decisions & Hypotheses

Log your open decisions and frame your active hypotheses.

/hypothesis-builder

Team Context

Map stakeholders and their concerns. Document cross-team dependencies and rituals.

/cross-team-mapper

Go Live

Run your first sprint kickoff. Test the Manager 1:1 playbook with real data. Add PM voice calibration once your communication-style section is filled or you have real writing samples. You're operational.

/playbook-sprint-kickoff
A note on examples

Screenshots and sample Claude responses in the course and docs are illustrative. The exact startup text or phrasing may vary by Claude Code version, but the workflow and save behavior should match the skill definitions in .claude/skills/.

Prerequisites
What you need installed
⚡

Claude Pro or Max

$20/mo or $100/mo subscription at claude.ai. Includes Claude Code access.

📚

Obsidian

Free download at obsidian.md. Your product knowledge vault lives here.

⌨

Node.js + Claude Code

Install Node.js LTS, then:
npm i -g @anthropic-ai/claude-code