A methodology execution engine for change managers
Change managers don’t lack methods. You already work with Cynefin, Theory of Change, Three Horizons, Wardley mapping. The hard part has never been the framework. It’s running it across a real group, and keeping what comes out of it.
The last few weeks of Harmonica updates are all in service of that. Sessions can now be chained into a full method: you pick the framework, Harmonica runs each stage as a facilitated session, carries context from one stage into the next, and collects the whole run into a single Project. Summaries point back at the participant’s exact words, so the synthesis is checkable. The facilitator can ask for a poll or a rating inline when a stage needs a number instead of a paragraph. A notification center tracks what’s moved. And creation now adapts its setup to the method you chose. The conversation was always there; what’s new is the engine that runs a whole method on top of it.
The engine also reads the room. Your organizational context (the HARMONICA.md you seeded at onboarding) shapes how every session runs by default. Per engagement, you can attach a research PDF, point at a previous session’s summary, or wire in an MCP tool the facilitator should consult mid-conversation. The facilitator pulls in what’s relevant where it’s relevant; nothing sits in a sidebar nobody opens.
Three things that engine is for. And one more, about how it gets sharper between runs.
1. Strategic planning, with Wardley mapping as the flagship
Run the Wardley chain and the discussion comes out as an actual map: components placed on the evolution axis, drawn from what the group said. It renders in Mermaid’s open-source Wardley syntax (v11.14.0+), which matters more than it sounds: the output is standard, portable text, not a picture trapped in our tool. You can correct the map directly when it gets something wrong, then take it anywhere Mermaid renders.
For a change manager, that closes the gap between the workshop and the artifact. A strategy conversation produces the diagram a strategy conversation is supposed to produce (the kind of thing you put in front of leadership) without a separate mapping exercise after the fact. More on the flagship method here: Wardley mapping with Harmonica.
2. Collective memory, so the work compounds
Most retrospectives are run as projects (a budget, a deliverable, a report nobody reads), which is why they’re the first thing cancelled when the calendar tightens. The cost of skipping them is invisible week to week and severe over years: the same debates return, context evaporates every time a cohort rotates out. What organizations actually want is closer to what brains do at night.
Chained retros are cheap enough to run every cycle, and their output is meant to accumulate: each one becoming atomic, linked notes the rest of the org can browse later, with a periodic pass that surfaces themes recurring across them. For a change manager, it’s the difference between client work that ends with a deck and client work that compounds.
3. Public sensemaking
The same engine runs outward, not just inside an org. When a city, a network, or a community needs to hear from hundreds of people and turn it into something decision-grade, Harmonica runs the interviews in parallel and synthesizes them into a dashboard and a browsable wiki: the Public Sensemaking Package. The synthesis lives at a public results URL with the participant quotes still attached; anyone with the link reads it directly, no login required. The output is a living artifact, not a screenshot at the end of an engagement. I made the longer case for this as civic infrastructure in a talk for OpenCivics earlier this year: AI-facilitated sensemaking as civic infrastructure. Strategy, memory, and sensemaking are the same machine pointed at three different rooms.
4. The engine improves with you
A method works as well as the facilitator running it. The new Review tab on the results page gives you an AI critique of how the facilitator actually behaved during this session, like where it pushed too long on one question or where it skipped a clean handoff, with each finding backed by the participant’s actual words. Where the review proposes a rule (something like “move on after two follow-ups on the same question”), you apply it in one click through the same Session Design confirm flow you’d use to edit the prompt by hand. The next time you run the same method, it runs with the rule in it. The engine doesn’t ask you to trust it more; it gives you something to sharpen between runs.
A direct ask
Harmonica is independent and we intend to keep it that way, which means funding this next stretch of work deliberately rather than raising. The most useful thing you can do, if any of the above is something you’d actually run, is buy a lifetime deal: $499 once, lifetime access, and it directly funds the runway to finish what’s described here. If you’re weighing it, now is the moment it helps most.
Pick a method and run it at app.harmonica.chat.