Mitchell Agentic Sprint

AI-led 6-step sprint that walks an AI builder through validating one idea — vertical, buyers, framework, positioning — to a sales deck, outreach plan, and pre-seed investor deck. Adversarial by default. Output is artifacts, not product.

View project on GitHub

ADR-009: Markdown-First Artifacts; Defer Slide Rendering to v0.2.0

Field Value
Decision ID ADR-009
Initiative Step 6 synthesis output format
Proposed By Chris Barlow
Date 2026-05-09
Status Accepted

WH(Y) Decision Statement

In the context of Step 6 producing the three concept-§4.6 artefacts (sales deck, customer outreach plan, pre-seed investor deck) for the user to take into customer conversations and pre-seed pitches,

facing the choice of artefact output format — markdown templates the user pastes into their own slide tool, generated .pptx files via python-pptx, generated .pdf files via Reveal.js + headless Chrome, or a polished slide-builder UI with branding controls,

we decided for markdown-first artefacts — three template files (sales-deck.md, outreach-plan.md, investor-deck.md) under docs/templates/, each with slide-per-section structure and speaker-notes blocks, populated by Step 6 with the user’s actual evidence from Steps 1–5,

and neglected PPTX generation in v0.1.0 (~1 week of python-pptx integration plus template-fidelity edge cases; pulls Extensive-tier scope into Moderate per ADR-002), PDF rendering via Reveal.js (additional toolchain — headless Chrome — for marginal-over-markdown gain), and a slide-builder UI (out of scope for a Claude Code plugin v0.1.0; designers will always want more controls),

to achieve the concept’s headline value at v0.1.0 (a builder walks out of Step 6 with three artefacts they can take to a customer conversation or a VC meeting this week), low rendering complexity (markdown survives clipboard, version control, code review, and email), preserved evidence chain (every slide cites the buyer interview / framework citation / competitor verbatim that supports it — slides without evidence are explicitly marked, not padded),

accepting that the user must paste the markdown into their preferred slide tool (Keynote / Google Slides / PowerPoint / Pitch / Marp) themselves, the artefacts will not have visual polish in v0.1.0 (acceptable — the concept frames v1 as “everything you need to brief your AI builder of choice” rather than a polished deliverable), and slide rendering becomes a v0.2.0 priority only if tester demand pulls it forward (per the recommendation’s “no Extensive without demand signal” rule).

Options Considered

Option Verdict
Markdown templates with slide-per-section structure (Selected) Highest leverage for v0.1.0; preserves evidence chain; user owns the rendering choice
python-pptx PPTX generation Rejected for v0.1.0 — ~1 week of integration + template fidelity edge cases; defer to v0.2.0
Reveal.js + headless Chrome PDF Rejected — additional toolchain for marginal-over-markdown gain
Custom slide-builder UI Rejected — out of scope for a Claude Code plugin

Dependencies

Relationship ADR ID Title Notes
Implements ADR-002 Build tier — Moderate Markdown-first is a Moderate-tier choice
Refines ADR-006 NotebookLM via MCP NotebookLM augments synthesis via prompt builders; output remains markdown

References

Reference ID Title Location
SPEC-009-A Artifact Template Format specs/SPEC-009-A-Artifact-Templates.md
Concept Mitchell Agentic Sprint concept §4.6 docs/concept/concept.md
Research Build options §1 / §2 risk: “investor deck weakness” mitigation docs/research/build-options.md

Status History

Status Approver Date
Accepted Chris Barlow 2026-05-09