Turn the AI agent you already use into a PM co-pilot that knows your product
13 battle-tested commands for Claude, Gemini, or Codex that write your PRDs, user stories, discovery hypotheses, and exec updates — grounded in your company's context, not generic templates. Set up once in 5 minutes.
Get the toolkit — $97 one-time Instant download · Lifetime updates · 30-day money-back guaranteeYou became a PM to ship products. Not to spend your week formatting documents.
Be honest about where your hours actually go:
- Another PRD written from a blank page, again — when the thinking was done days ago
- User stories rewritten three times because the acceptance criteria weren't testable
- A leadership update due Friday that swallows your Thursday evening
- ChatGPT drafts that sound impressive and say nothing — because it doesn't know your product, your users, or your metrics
- That quiet pressure in every planning meeting: "PMs who master AI will replace PMs who don't."
Here's the thing: the AI agents engineers use every day — Claude Code, Gemini CLI, Codex — are dramatically more capable than a chat window. They read files, follow structured workflows, and keep context. They're just not set up for PM work.
This toolkit sets them up for PM work.
From rough idea to Jira-ready tickets — in one sitting
This is the Delivery workflow. Type a command, answer a few sharp questions, get a structured artifact your engineers will actually read.
› /initiative let sellers offer subscription bundles with a free trial period Reading context/base-context.md, context/product-context.md… Two gaps before I draft: 1) Which seller segment is this for? 2) Is there a revenue target attached? › mid-size creators, target is +15% LTV by Q3 ✓ output/initiative-2026-06-09-subscription-bundles.md 8 sections · problem framing · success metrics · risks › /prd output/initiative-2026-06-09-subscription-bundles.md ✓ output/prd-2026-06-09-subscription-bundles.md — lean, engineer-focused › /user-story output/prd-2026-06-09-subscription-bundles.md ✓ 7 Jira-ready stories with testable acceptance criteria › /publish output/user-stories-2026-06-09-subscription-bundles.md ✓ Pushed to Jira — 7 stories created under EPIC-214
Notice what it did not do: it didn't invent a seller segment or hallucinate a metric. The toolkit's smart mode assesses what you gave it, asks about the gaps, and only writes what's true.
13 commands covering the whole PM job
Not a prompt dump — a system. Three connected workflows plus the context engine that grounds everything in your product.
🚚 Delivery — idea to shipped
/initiative | Strategic initiative doc: what, why, for whom — 8 sections |
/prd | Lean, engineer-focused PRD that skips the fluff |
/user-story | Jira-ready stories with testable acceptance criteria |
/epic | Jira epic with risk mapping (Marty Cagan's four big risks) |
/publish | Push artifacts straight to Jira or Confluence |
🔍 Discovery — evidence before code
/hypothesis-refine | Spars with you until vague ideas become falsifiable hypotheses |
/interview-analysis | Turns interview notes into insights and hypothesis verdicts |
/ost | Opportunity Solution Trees (Teresa Torres methodology) |
📣 Communication & quality
/decision-brief | Leadership decision docs — it pressure-tests your reasoning first, then simulates exec pushback |
/leadership-update | Exec reports that cross-reference last cycle's commitments |
/tracking | Event tracking specs that follow your analytics taxonomy |
/test-scenarios | Categorized QA test tables from any PRD or story |
/init | The 5-minute interactive setup that powers everything above |
Agent-native setup: your AI installs it for you
No installer, no account, no new app to learn. The toolkit is taught to your agent by your agent.
- Download & unzip
You get the full toolkit as a folder of readable markdown. No black boxes — you can open and edit every prompt.
- Open your agent in the folder
Claude Code, Gemini CLI, Codex — whichever you already use.
- Paste one prompt
The included Quickstart prompt makes your agent read the toolkit, interview you about your role, company and product, and generate your context files. From then on, every command knows your world.
That's the part most "AI for PM" resources get wrong: they hand you 200 prompts to copy-paste forever. This is a system your agent runs — and your context compounds with every artifact.
Why not just… prompt ChatGPT?
| Raw ChatGPT | PM course | Agentic PM Toolkit | |
|---|---|---|---|
| Knows your product & users | ✗ starts from zero every chat | ✗ theory only | ✓ context files, set up once |
| Output quality | Generic, needs heavy editing | You still write everything | ✓ structured artifacts, PM frameworks built in |
| Invents facts | Often | — | ✓ smart mode asks instead of inventing |
| Connects to Jira/Confluence | ✗ | ✗ | ✓ /publish command |
| Time to value | Hours of prompt fiddling | Weeks of lessons | ✓ 5-minute setup, first artifact today |
| Price | $20/month forever | $300–$2,000 | $97 once |
Built by a PM who ships with this every week
Aíquis Rodrigues is a Group Product Manager at Hotmart, one of Latin America's leading creator-economy platforms, with an MSc in Computer Science and 8+ years building digital products.
He co-hosts Papo na Arena, a top-30 tech podcast in Brazil about digital products and technology. This toolkit isn't a side experiment — it's the system he uses daily for his own initiatives, PRDs and leadership updates, refined in production and now packaged for you.
One payment. Yours forever.
Checkout localizes automatically: $97 · €97 · R$97
- All 13 commands — the complete prompt library, editable markdown
- Context system + interactive agent-led setup (EN & PT-BR)
- 3 connected workflows: Delivery, Discovery, Quality
- Jira & Confluence publishing via
/publish - Output templates for every artifact type
- Works with Claude Code, Gemini CLI, Codex, Antigravity
- All future toolkit updates included — no subscription
30-day "it pays for itself" guarantee. Use the toolkit for a month. If it doesn't save you at least one afternoon of document drudgery, reply to your receipt and you get every cent back. No forms, no questions.
Questions PMs actually ask
Which AI agents does it work with?
Claude Code, Gemini CLI, OpenAI Codex, and Google Antigravity — with native command wrappers for each. The core prompts are agent-agnostic markdown, so any capable agent that reads files will run them.
I'm a PM, not an engineer. Do I need to know the terminal?
If you can paste a prompt and answer questions in plain language, you're qualified. The Quickstart is literally: open the folder, paste one prompt, talk to your agent. No code, no configuration files to write by hand.
Why wouldn't I just use ChatGPT with my own prompts?
You can — that's what most PMs do, and it's why most AI output reads like a press release. Generic models don't know your product, your users or your metrics, and they confidently invent what they don't know. The toolkit fixes both: a context system that grounds every artifact in your world, and smart modes that ask about gaps instead of hallucinating them.
Does it work for my seniority level?
Yes, by design. Junior PMs get the frameworks of senior practitioners baked into every artifact (Cagan's risks, Teresa Torres' OST, falsifiable hypotheses). Senior PMs and leaders get the sparring modes — /decision-brief challenges weak reasoning and simulates exec pushback before you walk into the room.
Is it in English or Portuguese?
Both — and neither. The prompts auto-detect your language: write in Portuguese, get artifacts in Portuguese. Quickstart guides are included in English and Brazilian Portuguese.
What exactly do I receive after paying?
An instant download (plus a backup link in your receipt email): the complete toolkit folder — 13 commands, prompt library, templates, context system, examples and quickstart guides — as readable, editable markdown. All future updates included.
Is this a course?
No. There's nothing to watch and no modules to complete. It's a working system you install in 5 minutes and use on your real backlog the same day.
Can my whole team use one license?
One license = one PM. Artifacts you generate are yours forever and freely shareable, but the toolkit itself isn't. For team licensing, reach out on LinkedIn.
What if it's not for me?
30-day money-back guarantee. Reply to your purchase receipt and you're refunded in full — no questions, no friction.
The PMs who master agents aren't smarter. They started earlier.
For less than the cost of a team lunch, your next PRD writes itself around your context — and every artifact after that gets sharper.
Get the Agentic PM Toolkit — $97 Instant download · Lifetime updates · 30-day guarantee