1  # Pi-OpportunityMindset
  2  
  3  > *Mindset is the first thing that matters and the last thing anyone works
  4  > on. Sooner or later people realize that everything comes down to mindset.*
  5  
  6  A design project exploring one concrete question:
  7  
  8  **How do we exploit opportunity-mindset in artifacts we actually build —
  9  specifically agent personas and project-planning documents — where the
 10  "mindset" embedded in the work is directly editable?**
 11  
 12  This repository is research and design material, not a software library.
 13  The primary artifact today is [`design.org`](./design.org), a working
 14  document that develops the idea from first premise to a refined planning
 15  template with a worked example.
 16  
 17  ---
 18  
 19  ## Why This Project Exists
 20  
 21  Mindset is invisible to the toolchain. You can't `grep` it, refactor it, or
 22  unit-test it. Engineers reach for what their tools can hold, and mindset
 23  slips through. That structural neglect is what makes it a high-leverage
 24  place to intervene — *if* you can find artifacts where mindset is directly
 25  editable.
 26  
 27  Two such artifacts:
 28  
 29  - **Agent personas** — the agent's "mindset" is literally text in a system
 30    prompt. We can write the mindset into existence in a way we can't with
 31    humans.
 32  - **Planning documents** — the frame a plan opens with (risk, constraint,
 33    opportunity, outcome) shapes what the rest of the plan even attempts to
 34    capture.
 35  
 36  Both are places where a small edit to posture produces large downstream
 37  differences in what gets built.
 38  
 39  ## The Core Move: Pre-Parade Planning
 40  
 41  Standard planning runs *goal → tasks → estimates → risks*. The risk
 42  register does most of the work, and it is structurally pessimistic.
 43  
 44  Pre-parade planning inverts it. Instead of a **pre-mortem** ("imagine the
 45  project failed; what killed it?"), run a **pre-parade**:
 46  
 47  > *Imagine the project succeeded brilliantly. Work backwards and ask:
 48  > "what had to be true?"*
 49  
 50  The structural difference matters more than the attitude:
 51  
 52  - Pre-mortem produces a **risk list** — unordered, defensive.
 53  - Pre-parade produces a **causal chain from the goal backward** — ordered,
 54    generative.
 55  
 56  Same content surface area, different shape. Shape determines what you can
 57  do with it next.
 58  
 59  ## Current Shape of the Artifact
 60  
 61  After one round of audit, the pre-parade template settles into
 62  **four deliverables** and **five operations**:
 63  
 64  **Deliverables** (what lives on the page)
 65  
 66  1. **Parade scene** — actor, horizon, vivid concrete description of
 67     success, with the minimum celebrable version nested inside.
 68  2. **Required state, dependency-ordered** — facts about the world, not
 69     tasks; gaps in articulation surface as research items.
 70  3. **Bets and how we'd know they failed** — uncertainties paired with
 71     their disconfirmation signals by construction.
 72  4. **Cost cap** — what we will not spend, even for the parade. This is
 73     where opportunity-mindset is rescued from naive optimism.
 74  
 75  **Operations** (named checks applied while writing)
 76  
 77  - **Sharpness check** — write the one-sentence headline.
 78  - **Reduction check** — what's the smallest celebrable version?
 79  - **Actor check** — who is the protagonist?
 80  - **Horizon check** — by when?
 81  - **Opacity check** — where can't you state the next "had to be true"
 82    sharply? That's a research item, not a task item.
 83  
 84  The operations being *named* is the point. Users don't fill in boxes; they
 85  apply checks. That is the difference between a form and a practice.
 86  
 87  A worked example (a chapter of *The Little Schemer*) is included in the
 88  design doc to show the template under load.
 89  
 90  ## Goals
 91  
 92  - Develop pre-parade planning into something usable across personal
 93    learning, features, and refactors — not just a conceptual sketch.
 94  - Design the agent-persona counterpart: an **opportunity-first
 95    system-prompt skeleton** that preserves constraints as load-bearing for
 96    an opportunity rather than as free-standing fences.
 97  - Keep both artifacts audit-friendly: every section has to earn its place
 98    by answering *"if removed, what is actually lost?"*
 99  
100  ## Repository Contents
101  
102  | File | Purpose |
103  | --- | --- |
104  | [`design.org`](./design.org) | Primary design document. Premise, two angles, pre-parade template, audit, worked example, open questions. |
105  | `README.md` | This file. |
106  | `LICENSE` | GNU General Public License v3.0. |
107  
108  ## Future Improvements
109  
110  Open questions carried forward from the design document:
111  
112  - **Discoveries-during-planning slot** — the most valuable output of a
113    pre-parade is sometimes what you didn't know until writing forced you to
114    notice. Decide whether this gets its own section, nests in the parade
115    scene, or only appears in team versions.
116  - **Template vs. conversation** — maybe the artifact is not a template at
117    all but a set of questions an agent asks during conversation. Forms
118    make you fill boxes; conversations adapt to the shape of what you don't
119    yet know. Likely answer: form for solo use, conversation for kickoff.
120  - **Self-audit step** — should applying the template surface its own
121    weaknesses? ("After writing this, what would have to change for it to
122    still be useful in two weeks?") Worth adding if the artifact is meant
123    to live past initial planning.
124  - **Structural-vs-mindset trap** — "everything comes down to mindset" is
125    also what teams say to avoid fixing broken process. Opportunity-mindset
126    itself needs an obstacle-check. Working formulation:
127    *mindset is what you have left to work on once structure is healthy,
128    and what you can't substitute for when it isn't.*
129  
130  Possible next moves:
131  
132  - Apply the refined template to a real upcoming task and capture where it
133    strains.
134  - Draft the opportunity-first agent-persona skeleton and audit it the
135    same way the planning template was audited.
136  - Compare against existing planning skills (`/gmsd:plan-phase`,
137    `prd-to-plan`, `write-a-prd`) and identify which sections shift under
138    opportunity-first framing.
139  - Decide whether this becomes an installable **skill** or stays as
140    personal design material.
141  
142  ## Contributing
143  
144  This is early design work. The most useful contributions right now are:
145  
146  - Applying the pre-parade template to a real project and reporting where
147    it strained or misfired.
148  - Counter-examples — cases where obstacle-first framing was demonstrably
149    the better tool.
150  - Drafts of the agent-persona counterpart.
151  
152  Open an issue or a PR against `design.org`.
153  
154  ## License
155  
156  This project is licensed under the **GNU General Public License v3.0**.
157  See [`LICENSE`](./LICENSE) for the full text.
158  
159  The intent of choosing GPLv3 here is share-alike: templates, agent
160  personas, or tools derived from this work should remain open under the
161  same terms so the community can keep auditing and refining them in the
162  open.