settings — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited settings (Hook) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
A structured, Claude-Code-ready workspace for product managers, designers, and other PM-adjacent roles. Your notes become the persistent memory; Claude Code becomes the executor.
This is a template. Clone it, run /personalize --quick (three minutes), and you have a working AI workspace.
v2 is live (see CHANGELOG.md). Tasks now live in-repo atDashboard/tasks.md- no external task manager required. The old Todoist-coupled v1 is available on thev1tag if you need it.
Background reading: - Your AI Has No Memory. Mine Does. - the architecture that inspired this template. - The Four Claude Code Building Blocks - CLAUDE.md, MCP, skills, agents. - Claude Has 5 Memory Layers - how Claude Code actually remembers you.
Three steps. Two of them are skills.
1. Clone the template. Pick the path that matches how you work:
gh repo create my-pm-workspace --private --template aleksander-dytko/ai-pm-workspace && gh repo clone my-pm-workspace && cd my-pm-workspaceFull instructions with screenshots and gotchas: Quick start below and setup/SETUP-GUIDE.md.
2. Run `/personalize --quick` (3 minutes).
Open Claude Code in the cloned folder and type /personalize --quick. It asks three multiple-choice questions and one free-text question ("your name, your company in one sentence, your focus area"). Fills your identity into CLAUDE.md, trims the workspace to your role and audiences. You're ready to work.
3. Run `/guide` (eight modules, one per sitting).
/guide is an interactive learning path that walks you through every shipped skill using sample data that's already in the repo. No scrambling for a real meeting or a real decision to practice on. Role-adaptive so it reframes examples for designers, analysts, and engineers too.
After that, try the skills on real work. When you want more context on board, run /personalize --deep (10 min) to add MCP detection, active initiatives, and people profiles.
Dashboard/tasks.md - no Todoist or external tool required.+-------------------+ +---------------------+ +-------------------+
| | | | | |
| YOUR VAULT |<-->| CLAUDE CODE |<-->| MCP BRIDGES |
| (persistent | | (executor) | | (optional) |
| memory) | | | | |
| | | Skill categories: | | - GitHub |
| - Daily journal | | * Core daily | | - Your docs |
| - Meeting notes | | * Craft | | - [Your tools] |
| - Decisions | | * Onboarding | | |
| - Epics | | * Feedback | | |
| - Research | | | | |
| - Dashboard/ | | | | |
| tasks.md | | | | |
+-------------------+ +---------------------+ +-------------------+
^ |
| writes back |
+-------------------------+| Skill | What it does | |
|---|---|---|
| `/today [morning\ | evening]` | Morning focus planning or evening close-out. Morning fills journal and picks max 3 focus items; evening logs what got done and prepares tomorrow. |
/meeting | Turns raw meeting notes into a structured note with decisions, action items, and follow-up tasks for Dashboard/tasks.md. | |
/decision [topic] | Gathers context, drafts 2-3 options with trade-offs, creates a decision note, drafts a communication. | |
/communicate [topic + audience + channel] | Drafts Slack, email, or async doc matched to the audience. Uses context from recent meetings and decisions. | |
/weekly-plan | Weekly P-Tasks planning with an overplanning challenge and triage of Dashboard/tasks.md. |
| Skill | What it does |
|---|---|
/create-epic [idea] | Turns a raw idea into a structured epic draft using templates/epic-template.md. Sets stage to Explore. |
/user-journey [flow] | Persona-grounded journey map using research/personas/. Lands in research/journeys/. |
/competitive-research [topic] | Produces a sourced competitive matrix. Every observation cites a source or is marked inferred. |
/opportunity-solution-tree [outcome] | Teresa Torres discovery framework. Ported from phuryn/pm-skills under MIT; see the skill's ATTRIBUTION.md. |
| Skill | What it does | ||
|---|---|---|---|
| `/personalize [quick\ | deep]` | Quick (3 min) or deep (10 min) setup. Quick fills identity; deep adds MCP detection, initiatives, and people profiles. | |
| `/guide [module\ | next\ | reset]` | Interactive 8-module learning path through all the shipped skills. Role-adaptive. Uses sample data in samples/. |
| Skill | What it does |
|---|---|
/report [description] | Files a bug / confusion / feature request / feedback as a GitHub issue on this template repo. Uses GitHub MCP if available; falls back to a pre-filled issue URL. |
Pick the path that matches how you work.
#### Path 1: GitHub UI + GitHub Desktop (no terminal required)
Best for: designers, PMs, anyone who doesn't live in a terminal.
my-pm-workspace), make it private, create.File > Clone repository > [your new repo]./personalize. Done.If you get stuck on the GitHub side of this, Hannah Stulberg's Tool School: GitHub 101 is a great primer for non-developers, and setup/SETUP-GUIDE.md has platform-specific install commands.
#### Path 2: Ask Claude Code to do it (minimal terminal)
Best for: PMs who have Claude Code installed but aren't fluent with git.
Open Claude Code in any empty folder and paste:
Clone the template at https://github.com/aleksander-dytko/ai-pm-workspace
into a new repo called my-pm-workspace on my GitHub account,
then clone that new repo locally and open it.Claude will use the GitHub CLI to create the repo and clone it. If you don't have gh installed, Claude will guide you through installing it.
Then run /personalize.
#### Path 3: Git CLI (classic)
Best for: developers.
gh repo create my-pm-workspace --private --template aleksander-dytko/ai-pm-workspace
gh repo clone my-pm-workspace
cd my-pm-workspace
claude # run Claude CodeThen run /personalize.
/personalize doesCLAUDE.md identity. Ships working skills immediately./personalize --deep when you want more context.Once /personalize --quick is done, you have two paths.
Try the skills one by one:
/today - morning planning or evening close-out./meeting - process a raw meeting note into a structured one./decision - document a product decision with full context./communicate - draft a Slack, email, or async update./weekly-plan - Sunday/Monday planning with an overplanning challenge.Or run the interactive guide:
/guideEight modules, role-adaptive, using sample data shipped with the template. Covers all shipped skills plus how to extend your workspace with external skills. Good to do over a week, one module per day.
When you want more, /personalize --deep adds MCP integration, initiatives, and people profiles.
CLAUDE.md is the single most important file in this setup. Claude Code reads it at the start of every session, so it always knows:
The more context you put into CLAUDE.md, the less you need to explain in each conversation. Deep dive: Claude Has 5 Memory Layers.
Skills are Markdown files in .claude/skills/[name]/SKILL.md. Each has:
name and description (how Claude discovers the skill).Ideas to build yourself:
/retro - sprint or quarterly retrospective from journals and meeting notes./standup - daily standup summary from journal + today's plan./health - vault health audit (unlinked notes, stale tasks, orphaned decisions).Module 8 of /guide walks you through how to write your own or adopt one from the internet.
There's a growing open-source library of PM skills you can borrow.
opportunity-solution-tree) already ships with this template as a worked attribution example.To add another phuryn skill to your workspace, copy its folder into .claude/skills/, keep the attribution header, restart Claude Code.
This template keeps tasks in-repo (Dashboard/tasks.md) so you can start without any external tool. If you want Todoist sync later, see docs/todoist-sync.md.
Three principles:
If you cloned v1 (Todoist-coupled) and want to upgrade: see the migration notes in CHANGELOG.md. Short version: tasks now live in Dashboard/tasks.md instead of Todoist, and four skills (/end-of-day, /process-inbox, /slack-thread, and parts of /personalize) were either merged or removed. Your notes and journals are unchanged.
This is an open template. If you build skills, patterns, or integrations that others could use, PRs are welcome.
Bugs, questions, or "I got stuck here" feedback: open an issue or run /report from within your workspace.
MIT - see LICENSE.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.