The manual, not the framework. Take what you need.

No to either. There is no sign-up, no gate and no fee. Tenhaw publishes all seventeen guides in full, with 142 steps written out rather than summarised, while the paid version is the Programme and Delivery Management engagement that installs the same method alongside a client's own teams. Each guide carries a worked example, the failure modes and a definition of done. Copy them, adapt them and rename them to fit the words your organisation already uses, and you owe nothing for doing it. None of that is an accident. Tenhaw engagements are built to leave, and a method a team can only run while it is paying for it is not one they can be left with.

The product manager, delivery lead or engineer who is running a quarter and doing the work the operating model asks for, from pricing an outcome to running a team health check. The guides themselves are operating instructions rather than a sales case, so a buyer evaluating the firm should start with the engagements or the case studies instead, because that is where the agentic work and the evidence for it live. Tenhaw publishes them free and prices the installed version separately, five engagements staffed as delivery teams of two or three senior people, with James Rooney leading the audit personally and providing oversight on the rest.

The full life of a quarter, in five stages that follow the order work moves through The Tenhaw Way. Setting up the work covers pricing outcomes, building a quarterly roadmap, writing value-carrying epics, breaking them into stories and running discovery. Writing the work covers stories, chapters, bug reports and the RAID log. Shipping and supporting covers release notes, the first week after a release and root cause analysis. Measuring and managing covers value validation, running delivery day to day and delivering on time. Forecasting and team health covers confidence-interval forecasting and the monthly team health check.

The Tenhaw Way is the operating model in one read: four values, two delivery modes, a quarterly timebox and seven rituals. The how-to guides are the manual underneath it. Each of the seventeen moves the model asks you to make is written out with the steps, a worked example and the ways it goes wrong. Read the model to decide whether you want to run this way. Open a guide when it is Monday morning and you actually have to make one of the moves.

Because delivery is the method underneath the agentic work, not a separate business. Tenhaw installs the same operating model alongside client teams as Programme and Delivery Management, bought on its own with no requirement that Tenhaw builds anything. Priced outcomes, quarterly roadmaps and monthly value validation are what turn an agentic build into money the business can see arriving. Those engagements report monthly against a value commitment, and a month that delivers no measurable value is reported as a failed month. An agentic system with no loop between built and worked never shows up in the accounts. The agentic work itself lives in those engagements and the case studies, and the seventeen free how-to guides are the discipline that makes it pay.

No, take what you need. The library is written as a manual, not a framework. Each guide stands alone with its own steps, worked example, failure modes and definition of done, so a team can adopt the bug report format or the release notes structure without touching anything else. The spine the guides do share is that everything traces to a priced outcome, so the further up the stack you go, the more the pieces reinforce each other. Start with the one your week actually needs and let the rest earn their way in.

Yes, they are free to copy, adapt and rename. If your organisation says initiative instead of outcome, or workstream instead of epic, keep your words and take the mechanics: the priced target, the visible arithmetic, the named owner, the monthly validation. What matters is the discipline each guide encodes, not the vocabulary it happens to use. The one renaming worth resisting is any that quietly removes the number, because a level of work that no longer carries currency has stopped being the same thing.

Two of the seventeen, how to measure the value a project delivered and how to forecast delivery with confidence intervals. The other fifteen change how the work is written. These two change what leaves the delivery organisation. Did the money arrive, and when will the rest of it land? Value validation runs monthly on every live outcome, and it is the ritual most organisations skip, which is why so few can answer what the quarter actually produced. Tenhaw runs both alongside client teams on a Programme and Delivery Management engagement, where every commitment carries a p50 and a p85 date simulated from the team's own throughput. If your board is asking either question, start with those two.

Every guide carries an AI-native note saying exactly what changes, and the arithmetic never does, because outcome pricing, epic value shares and the capacity maths read the same in both modes. What changes sits below the epic. AI-native teams stop there, so the outcome ticket itself carries the currency share, the key user journeys and the test requirements, and the three guides on stories and chapters do not apply at all. Gating and forecasting shift with it. An epic is approved on journeys and test requirements rather than an attached story, and throughput is sampled over fewer, larger items.

Operators who install the operating model alongside your teams and stay until it sticks, on published prices: a four-week AI Readiness Audit at £44,000 fixed, a two-person Agentic Design Team at £35,000 to £55,000 a month, up to a three-person Agentic Build Team at £70,000 to £85,000 a month. Reading a guide and running a quarter on it are different work, which is why an engagement brings people who have run these rituals before, pairs them with your team, and leaves capability behind rather than dependency. The same operating model sits underneath the agentic engagements, where it is what turns a build into validated value. The seventeen guides stay free either way.

Both. Every guide sets its steps out in full, then works the same move end to end on a named example, eighteen of them across the seventeen guides. Those organisations and their figures are invented so the shape can be shown whole, while Tenhaw's real work sits in named case studies for Anglo American, Greggs, Yondr and Colart. Pricing an outcome carries two examples, one on a revenue line at an invented retailer and one on an operational cost base at an invented insurer, since a conversion rate teaches you nothing about a claims operation. Each runs the arithmetic from the baseline and the date it was read through to the condition that would make you call the outcome not realised.

Each guide prints its own run time on the index, next to the number of steps, so you can see the cost before you open it. They range from about ninety minutes to raise and triage a bug, through six hours to price an outcome across a few sittings, to about two days for a quarterly roadmap written two or three weeks before the quarter starts. The recurring rituals are lighter than teams expect, with value validation about three hours a month and the team health check about two, each built round a forty-five minute session. Discovery is capped at two weeks, and the default post-release monitoring window is seven days.

No, that is the other half of the method. This library is the delivery side, covering how work gets priced, planned, written, shipped, measured and forecast. The engineering side is Tenhaw's Building with AI, which treats requirements as the source code, finds the gaps in them before any code exists, and enforces an open-source standard of 72 rules with RFC 2119 severities from the first prompt. The two meet at the ticket. These seventeen guides take a business result down to an approved epic and back up to whether the money arrived, and Building with AI takes that epic into the codebase. Both are published in full and free to use.

For some of them, yes, and not many people beyond that. Pricing an outcome needs whoever owns the reporting system sitting with you while the baseline is pulled, finance's own margin and lifetime value figures rather than your estimate of them, and one named owner senior enough to carry the number in next year's plan. Getting an epic into the quarter needs a product approval and an engineering approval, which is why that guide runs to half a day rather than an hour. The monthly health check needs the whole team scoring privately, and management reading the card unedited the same day.