Tools · Publishing
Guide
Turn a CanonicAI capability corpus into a published, cited how-to guide keyed to a job profile.
How it works
Composed grounding, never ad-hoc retrieval: the cross-book reconciled cluster_model is the spine (core constructs → sections, relationships → the founder sequence, open_divergences → honest tensions); per-book extractions add frameworks + LibraryId-backlinked citations. Every claim cites the corpus by book_id; the corpus's real disagreements are surfaced, not flattened; provenance enforced in code; status always draft.
You bring
{ cluster?: 'start-a-company', profile?, mode?: aspire|resolve, audience? }
You get
CapabilityGuide (title · overview · sequence · sections[how_to/watch_out_for/grounded_in] · tensions)
Use it for
- →Aspire-mode on-ramp: 'How to Start a Company' guide for the founder profile, grounded in the entrepreneurship corpus
- →Attach a cited capability guide to a jobframe family×level profile as a sellable value-ladder rung
- →Surface a corpus's genuine tensions (blitzscale vs. bootstrap; raise vs. self-fund) as decision guidance, not platitudes
See it work
example outputFocused guide "Fund Your Startup" carved from the start-a-company cluster, aspire mode, keyed to a first-time-founder profile.
Fund Your Startup
A first-time founder's guide to raising — or deliberately not raising — money
Aspire mode · for the founder who has a working product and a few customers and is deciding how to finance the next 18 months. This is a focused sub-guide within the broader Start a Company canon: it covers only the funding constructs, sequenced the way the corpus says to face them.
What "good" looks like
- Foundations: you can state how much you need and what it buys, in months of runway.
- Practitioner: you can choose between bootstrapping and raising based on your market's capital intensity — not on what's fashionable.
- Advanced: you can run a process that creates real competition for your round and hold terms because you have a credible alternative (a profitable path).
The path
- Know your number — size the raise from the milestone it must reach, not a round-number ask.
- Bootstrap vs. raise — decide whether outside capital actually fits this business.
- Instrument the use of funds — tie the money to validated learning, not headcount.
Section: Bootstrap vs. raise (Practitioner)
Misconception → Reality. Myth: raising money is the milestone — funded means winning. Reality: venture capital is a tool with a required shape; it fits businesses that can plausibly return a fund, and it actively harms ones that can't.
How to: estimate capital intensity (how much cash before the business pays for itself); name your target outcome honestly (lifestyle, acquisition, or fund-returning scale); only then match the instrument.
Watch out for: raising because peers did; taking "smart money" that buys control you'll regret; using a round to paper over a product that hasn't found demand.
A real tension the corpus won't flatten
Blitzscale for dominance vs. deliberate, profitable growth. Contested. The right answer depends on whether your market has a winner-take-most prize. In a network-effects land grab, speed funded by capital can be rational; in a steady B2B niche, it's how you blow up a good business chasing a bad one.
Grounded in The Lean Startup (Ries), The Four Steps to the Epiphany (Blank), and Traction (Wickman). Every claim cites the corpus; status: draft.
Run it on your data
Call it on your own inputs — over the API, or hand it to your AI agent via MCP. Discovery is open; running it is metered.