BOOK TWO — COWORK SCANNING PROMPT v2
Surfacing the Substrate — Revised
Revised prompt for Claude in Cowork to scan Paul Roebuck’s My Documents folder for material relating to “All Behaviour Is Learned. Not All Behaviour Is Taught.” 24 May 2026.
What changed from v1
The v1 prompt treated the book as a standalone behaviour-development project. It scanned for material on the fifteen-step developmental sequence, on behaviour, learning, change.
The v2 prompt incorporates a structural observation surfaced after v1 was written: NGE-FOF is not adjacent to the book — it is the canonical example of the book’s central argument. The directional defences (Not Good Enough, Fear of Failure / Not My Fault) are themselves learned behaviours that were never taught. They surface across jobs, in Hunter–Farmer dynamics, in Pegs / Holes / Hammers, in relationships, parenting, performance, illness, grief.
That changes the scan. The substrate for the new book lives across a wider area of the archive than the v1 prompt was looking at. Material on NGE-FOF, on shame, on protection patterns, on Hunter–Farmer, on Pegs / Holes / Hammers, on directional defences — all of it now belongs in the inventory at high relevance, not at low relevance. The v2 prompt instructs Cowork accordingly.
Paste the prompt below into Cowork. The supporting notes after the prompt are for your reference, not for Cowork.
The prompt — paste this into Cowork
I am Paul Roebuck. I am writing a book provisionally titled “All Behaviour Is Learned. Not All Behaviour Is Taught.” The book is a general-readership behaviour-change book, written from thirty years of clinical, organisational, and personal experience. Its central claim is that all adult behaviour was acquired through learning, but very little of that learning was deliberately taught, and the change strategies popularised by the contemporary behaviour-change literature mistreat the difference.
The book is structured around a fifteen-step developmental sequence I have used and refined in workshops, executive coaching, training and clinical practice since 2017: Recognition, Perception, Ownership, Internal Freedom, Identity, Aspiration, Belief, Vision, Action, Progress over Perfection, Failure, Persistence, Containment, Expansion, Meaning.
The book is structurally underpinned by a clinical framework I have developed over thirty years: the Shame-Response Continuum, and within it the directional defence pair NGE (Not Good Enough) and FOF (Fear of Failure / Not My Fault). NGE and FOF are themselves the canonical examples of the book’s central argument — these directional protection patterns are learned, never taught, and they surface across adult life in jobs, in Hunter–Farmer dynamics, in Pegs / Holes / Hammers, in relationships, in performance, in illness and grief. Material on NGE-FOF and on the wider Shame-Response Continuum is therefore HIGH-RELEVANCE substrate for this book, not adjacent material. Treat it accordingly.
I have been doing this work in various forms since the late 1980s. Material relevant to the book will exist across my entire archive. Scan the My Documents folder on this Mac comprehensively, including all subfolders. Where you encounter folders containing large numbers of files of unclear relevance, scan them sufficiently to determine whether individual files relate to the book’s subject. Do not skip folders based on their names — relevant material may be stored under client folders, project folders, archived training, personal correspondence folders, or anywhere else.
Material relevant to this book includes, but is not limited to, the following categories:
(a) The 2017 master slide deck on the developmental sequence (Recognition through Meaning), and any descendant versions, revisions, or workshop adaptations of that deck.
(b) Mind Gym training materials and workshop content from the early 2000s — the period when I first worked at scale on behaviour and emotional intelligence training.
(c) Slide decks from executive coaching engagements 2008 to 2026; client briefings, board presentations, and pitch documents from commercial work.
(d) Therapy notes, case formulations, and reflective journals from clinical psychotherapy practice 2016 onward. Flag confidentiality concerns; do not surface client material in identifying detail — indicate that material exists and what theme it relates to, without reproducing identifying content.
(e) Blog posts, articles, and website drafts from paulroebuck.co.uk; talks, podcasts, transcripts and speech preparation notes.
(f) Material on NGE-FOF, the Shame-Response Continuum, Where Does The Shame Go, Defending the Gap, anti-disintegration adaptation, directional shame regulation, Pegs / Holes / Hammers, Hunter–Farmer dynamics, primary defence patterns. This material is HIGH RELEVANCE for the new book even though it relates to my other books — it is the substrate the new book sits on.
(g) Mind the Gap manuscript material (the AI book) — this is my prior book, currently being prepared for publication. Flag it separately so I can keep its territory distinct from the new book. The new book is the human-side companion; Mind the Gap is the AI-side companion. They share the framework but observe different objects.
(h) Letters, emails, LinkedIn posts, longer-form social writing where I have explained or articulated the framework, the developmental sequence, NGE-FOF, or any of the related concepts to colleagues, clients, family, or audiences.
(i) Book proposals or chapter drafts I have started and not finished — any prior attempts to write any of this material as a book, including outlines, opening chapters, abandoned manuscripts, agent correspondence, publisher feedback.
(j) Reading notes and annotations on other authors’ work — specifically: my notes on Bowlby, Winnicott, Kohut, Bion, Nathanson, attachment theory, object relations, shame literature, executive coaching literature, Mind Gym alumni publications, contemporary behaviour-change literature (Clear, Duhigg, Fogg, Dweck, Kahneman).
(k) Training session plans, coaching session preparation notes, debriefs after major engagements — the practical apparatus of how I have delivered this work in front of audiences and in coaching rooms.
(l) Personal reflective writing, journals, diary entries, where I have worked out my own thinking about behaviour, change, learning, identity, defence — the private substrate underneath the public material.
For each file you identify as potentially relevant, produce an inventory entry containing: (1) filename and full path; (2) file format and approximate size; (3) date created or last modified; (4) a one-to-three-sentence summary of what the file contains; (5) which of the fifteen developmental sequence stages the file most closely relates to — or whether it provides framework/scaffolding rather than stage-specific content; (6) which content category from the list above (a) through (l) the file belongs to; (7) a relevance grade from one to five, where one means “mentions relevant themes in passing” and five means “directly addresses the book’s central argument, the developmental sequence, or the framework substrate”; (8) a confidentiality flag for any client material, therapy notes, identifying personal correspondence, or commercially sensitive content.
Organise the inventory by relevance grade, with grade-five files first. Within each grade, group by content category. Within each category, order by date — oldest first — so that the developmental arc of my own thinking is visible across years.
Where files appear to be different versions of the same underlying document, group them together and indicate which is the most recent. The historical versions matter because they show the thinking evolving.
Produce, in addition to the per-file inventory, a separate summary section identifying:
(i) any file that appears to be the 2017 master slide deck or its descendants;
(ii) any file using the phrase “all behaviour is learned” or close variants;
(iii) the three to five files in the archive that, in your judgement, most concentrate the book’s central argument or its richest case material;
(iv) any cluster of files that constitutes a coherent body of work — for example: all Mind Gym training materials from a single period; all coaching engagement notes for a particular sector; all writing on a single framework concept;
(v) the framework footprint — specifically: how much material exists on NGE-FOF, how much on Hunter–Farmer, how much on Pegs / Holes / Hammers, how much on the wider Shame-Response Continuum. This footprint matters because the framework is the substrate underneath the book;
(vi) any gap that strikes you — a developmental stage from the sequence for which no substantial material seems to exist, or a framework concept that is referenced but not developed in any file. The gaps matter as much as the surfaces because they tell me what the book’s new writing will need to fill.
Do not move, copy, rename, or modify any file during the scan. The inventory is a reading, not a reorganisation. Where you make judgement calls about relevance or categorisation, name the judgement and the reasoning briefly so I can override.
Output the inventory as a single Word document, with the per-file entries followed by the summary section. Save it to my Desktop with today’s date in the filename. If the inventory grows too large for a single document, split by relevance grade but keep the summary section as a separate, smaller document I can read in one sitting.
Two final instructions. First: where you genuinely cannot determine whether a file is relevant, err on the side of including it at low relevance grade rather than excluding it. The cost of a false positive is small; the cost of a missed file may be significant. Second: where you find material that strikes you as important but does not fit anywhere in the categories above — a fragment, a slide, a case note, a piece of writing that strikes you as significant without obvious placement — capture it in a separate “Worth surfacing” section at the end of the summary document, with your reasoning. I trust your judgement on this.
What changed substantively from v1
Three additions matter most.
First. NGE-FOF and the Shame-Response Continuum are now named explicitly as high-relevance substrate. In v1, the prompt told Cowork to flag framework material as adjacent. In v2, the prompt tells Cowork that the framework is the substrate — not adjacent. This will likely double or triple the volume of grade-four and grade-five material the scan surfaces.
Second. The content categories now run to twelve items (a through l) rather than seven. The new categories pick up the framework material (f), the Mind the Gap separation (g), and the personal reflective writing (l) that v1 underweighted.
Third. The summary section now asks Cowork for a framework footprint — how much material exists on NGE-FOF, Hunter–Farmer, Pegs / Holes / Hammers, the wider Shame-Response Continuum. This is the strategic intelligence the scan needs to produce. It tells you not just what is there, but how much there is on what the book is built on.
Practical guidance for running the scan
Run the scan when you have time to let it complete. Depending on archive size and file count, the scan may take several hours. Cowork operates locally; nothing leaves your machine.
When the inventory is in hand, read the summary section first. Do not attempt the full per-file inventory in one sitting. The summary tells you what was found at the level of clusters, gaps, and high-value files. That is the strategic view.
Bring the summary back into a fresh Claude conversation when you are ready to do the mapping work. Share the summary and the high-value file titles — not the client material itself. The mapping conversation will produce the working chapter outline.
Estimated end-to-end timeline: scan one to several hours; summary review one focused hour; mapping conversation ninety minutes to two hours. Total from prompt-to-outline: probably less than a working day if you do it in one sitting.
Confidentiality — explicit
The prompt instructs Cowork to flag sensitive material. Client material, therapy notes, identifying correspondence, and commercially sensitive content all need flagging. Cowork operates locally; the inventory does not leave your machine unless you share it. When you bring the summary into a conversation here, share the summary and file titles rather than client content. Anonymised case material is brought in later, chapter by chapter, under your control.
Emotional substrate — explicit
The scan will surface eight years and more of your own thinking. Mind Gym material. Early client engagements. Training sessions that mattered. Correspondence from people no longer present. Drafts you set aside. This is real. Pace it. Strategic summary in one sitting; deep dives into specific files at a rhythm that suits you. The inventory is not just categorisation; it is, for you, an encounter with the archive of your own thirty years.
Provenance
This v2 prompt and the surrounding guidance were drafted by Claude (Anthropic, Claude Opus 4.7) at Paul Roebuck’s explicit request on 24 May 2026, after Paul surfaced the structural observation that NGE-FOF is the canonical example of the new book’s central argument rather than adjacent to it.
The prompt may be revised further as the scan reveals what kinds of material actually exist in the archive. First scan often surfaces patterns that suggest better second-scan prompts. Treat this prompt as v2, not as final.
Cowork scanning prompt v2 for Book Two. Compiled 24 May 2026.
Subject, archive, framework, final say: Paul Roebuck.
Prompt drafting and structure: Claude AI (Anthropic) — claude-opus-4-7 | 24 May 2026.