nanopm

Archive one entity (opportunity, solution, objective, competitor, or persona) with a polished, contextual reason sentence. Argument: <type>/<slug> (e.g. opportunity/cold-start-context or solution/inline-viewer-pane). User context: the raw reason and kind — 'implemented: ...', 'dropped: ...', 'absorbed: ...' for opportunities/solutions; 'completed: ...', 'abandoned: ...' for objectives; 'defunct: ...', 'pivoted: ...', 'irrelevant: ...' for competitors; 'outgrown: ...', 'dropped: ...' for personas.

Challenge Me. Adversarial product challenge: a skeptical-CPO read of what you're building, who for, and the biggest strategic gap — then three direct challenges you should answer, starting with the question you're avoiding. Produces .nanopm/wiki/docs/challenges.md.

Build and maintain a ranked database of user opportunities (Teresa Torres-style — the user problems behind what you build, not the solutions). Stored as an LLM-wiki under .nanopm/wiki/entities/opportunities/: one file per opportunity + a ranked INDEX, a LOG, and an editable SCHEMA. bootstrap drafts the initial set from feedback + your assumptions + Nano's hypotheses, each marked by provenance; add captures one problem at a time. Two levels only (Theme → Opportunity); no scoring at v1 — a coarse priority instead.

Configure Supabase cloud sync for nanopm. One-time setup per machine: enter your project URL + anon key, run the migration, push your local wiki to the cloud. After setup, wiki pages sync automatically on every skill run.

Define who you're building for. Reverse-engineers personas from the codebase and prior nanopm artifacts when they exist, or builds them from scratch by interviewing you when the repo is empty. Produces the wiki Define page `.nanopm/wiki/docs/personas.md` — JTBD proto-personas plus an explicit anti-persona.

Aggregate user feedback from Dovetail, Productboard, Notion, Linear, and GitHub. Cluster into themes, surface the top unaddressed signal, map to current roadmap. Produces the wiki Feedback page (.nanopm/wiki/docs/feedback.md) — the primary input for all downstream PM skills.

Quantitative data analysis for PMs. Answers a specific product question using PostHog or Amplitude — trends, funnels, retention, paths. Writes findings to the wiki data page at .nanopm/wiki/docs/data.md, consumed by /pm-challenge-me and /pm-prd.

PM retrospective. Compares the roadmap page's NOW items against actual commits since the last roadmap run. Surfaces what shipped, what drifted, and what to carry forward. Closes the planning loop.

Monitor competitor products for changes: changelogs, API docs, new endpoints, pricing, and product updates. Snapshots pages, diffs against last run, produces a structured intel report. Can discover new competitors and run a full SWOT + positioning analysis on demand.

Define product objectives (OKRs). Reads challenge-session context, asks 2-3 clarifying questions, produces the wiki objectives page at .nanopm/wiki/docs/objectives.md with measurable goals and key results.

Brainstorm a compared set of solutions for one opportunity. Dispatches three fixed expert lenses (Eng / Design / Business) concurrently to diverge, converges them into ≥3 framed solutions — each with a pitch, appetite, impact, riskiest assumption, cheapest test, originating lens, and a one-line dissent note — and writes them under .nanopm/wiki/entities/solutions/ as the Solutions node of the Opportunity Solution Tree. A recipe over the existing query → reasoning → ingest primitives. The founder shortlists and chooses; the agent never auto-chooses.

The single ingestion door for raw user signal. Accepts a file path, pasted text (--paste), or a connector source (Granola, Dovetail, Google Drive, tickets), archives it verbatim under .nanopm/raw/<type>/<id> before extraction, applies a Mom-Test discovery filter to pull problems + verbatim quotes, then runs the learning loop: Pass 1 grounds EXISTING opportunities (append verbatim, propose a provenance upgrade toward evidence-backed), Pass 2 surfaces NEW strategy-coherent opportunities. Every grounded/created opportunity is bidirectionally linked to its raw source, and the run ends with a machine-readable run-summary. One batched human confirmation before writes; headless (viewer) runs fall back to write-then-review with ⚠ low-confidence.

Jam with Nano, your expert CPO. An informal, context-loaded thinking partner for product ideas, user problems, and what-to-build-next — Nano knows your company context and current objectives, no gate, no PRD. Sessions are named and resumable via your host's native session resume.

Break a PRD into engineering tasks and hand off to one of six peer targets: Linear, GitHub Issues, OpenSpec, gstack, Symphony (OpenAI's orchestrator), or Human-readable markdown. No preferred default — you pick the target.

Product discovery. Figures out WHAT to build before you plan HOW to build it. Maps the opportunity space, surfaces the riskiest assumptions, and designs the cheapest tests. Run this before pm-challenge-me when you're pre-product, pivoting, or unsure if a feature is right.

Write a product spec for a specific feature. Adapts to your team's methodology: Shape Up pitch (problem, appetite, solution, rabbit holes, no-gos) or standard PRD.

Build product roadmap. Reads strategy + objectives + connector data, asks clarifying questions, produces a roadmap adapted to your team's methodology (Shape Up bets, Scrum sprints, or NOW/NEXT/LATER).

Define product strategy. Reads objectives + challenge-session context, generates a strategy, dispatches adversarial subagent to challenge it, synthesizes into the wiki Plan page (.nanopm/wiki/docs/strategy.md).

Define how the company makes money. Reverse-engineers the business model from pricing pages, prior nanopm artifacts, and the codebase when they exist, or builds it from scratch by interviewing you when the repo is empty. Produces the wiki Define page `.nanopm/wiki/docs/business-model.md` — model type, revenue streams, pricing & packaging, and GTM motion.

Map who's who and who decides what. Reverse-engineers the org from prior nanopm artifacts, git history, and the team/about page when they exist, or builds it from scratch by interviewing you when the repo is empty. Produces the wiki Define page `.nanopm/wiki/docs/org.md` — the org map, key roles, decision-makers, and ways of working.

Define why the company exists and where it's going. Reverse-engineers mission, vision, and values from prior nanopm artifacts and the public site when they exist, or builds them from scratch by interviewing you when the repo is empty. Produces the wiki Define page `.nanopm/wiki/docs/vision-mission.md` — mission, vision, core values, and company stage.

Daily standup briefing. Pulls commits from all your active GitHub repos, Linear, Google Calendar, and Granola. Works from a standalone Product OS folder — no need to be inside a codebase. Surfaces today's meetings, cross-repo activity, priorities, and drift.

Map the product. Reverse-engineers what the product actually is — surface area, features, core workflows, technical bets — from the codebase and the public site when they exist, or defines the product concept from scratch by interviewing you when there's nothing built yet. Produces the wiki Define page `.nanopm/wiki/docs/product.md`. Absorbs the old pm-scan.

Weekly stakeholder update. Drafts a clear, honest status email for your manager, CEO, investors, or team. Reads what shipped, what slipped, and what changed strategically. Adapts tone to the audience.

Run the full nanopm PM pipeline: feedback → personas → challenges → objectives → strategy → roadmap → PRD. One command for a complete planning cycle. Each skill reads the prior output — the pipeline compounds.

Upgrade nanopm to the latest version. Detects install type, runs the upgrade, shows what's new. Also handles inline upgrade prompts when UPGRADE_AVAILABLE appears in any skill preamble.