Build a customer-specific Scalar solution proposal: start from the customer's challenges (hearing notes, meeting minutes) and map them to ScalarDB / ScalarDL. Use for: 提案書を作って, 提案スライド, 顧客課題, ソリューション提案, 〈顧客名〉向けの提案資料. Not: introduction decks with no specific customer (scalar-product-slides); non-Scalar proposals (google-slides-template); PPTX authoring (document-skills:pptx).
Scanned 9/19/2026
Install to Claude Code
npx -y skills add wfukatsu/slide-forge --skill scalar-proposal-slides --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Scalar Proposal Slides?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/wfukatsu-scalar-proposal-slides)More formats (shields.io, HTML) on the badges page.
---
name: scalar-proposal-slides
description: >-
Build a customer-specific Scalar solution proposal: start from the
customer's challenges (hearing notes, meeting minutes) and map them to
ScalarDB / ScalarDL.
Use for: 提案書を作って, 提案スライド, 顧客課題, ソリューション提案,
〈顧客名〉向けの提案資料.
Not: introduction decks with no specific customer (scalar-product-slides);
non-Scalar proposals (google-slides-template); PPTX authoring
(document-skills:pptx).
---
*[日本語](SKILL.ja.md)*
# Scalar Solution Proposal Slides
Use `references/scalar/workflow-contract.md` for shared sales-material rules
and `references/scalar/research-policy.md` for freshness/research. This skill
owns only proposal-specific decisions and generation.
Working directory: the slide-forge root — `${CLAUDE_PLUGIN_ROOT}` when running
from an installed plugin, `/path/to/slide-forge` on a local clone
(literal `cd` paths below assume the local clone).
## Important
- **Prerequisite skill**: `google-slides-template` (same repo) — auth, the
shared venv, the `scalar-2026` template, drawing API, and QA tooling. This
skill owns only what is proposal-specific: the hearing checklist, the
challenge→product mapping, and the proposal deck structure.
- **The customer's challenges drive everything.** A proposal without agreed
challenges is a product intro — route it to `scalar-product-slides`. Collect
hearing material (minutes, notes, RFP) before designing slides; what is not
known must surface as "to confirm today" on the deck, never as a guess.
- **Never fabricate customer-specific numbers.** Quantified effects need a
calculation basis from the hearing; otherwise write qualitative effects and
route quantification to the PoC ("measure it during the PoC and use it as
material for internal approval"). Public case numbers (ENS approx. 1/5, etc.)
are usable with sources.
- **Every initial proposal includes an architecture diagram and a BOM.** The
standard topology is 3 environments — development (local) / test
(aidd-infra-test) / staging (aidd-infra-staging) — on **AWS by default**
(rebuild the same role split on GCP/Azure only if the customer specifies).
After composing the architecture, output the cloud-service list and the
Scalar product list with quantities (and monthly license cost when
quantities are not specified) — on the deck AND as a list in the final
report (proposal-map.md §6).
- **Check the constraints before proposing** —
`references/scalar/proposal-map.md` §4 (cases where ScalarDB/DL does not
fit). Do not force a challenge onto a Scalar product; saying so is part of
proposal quality.
- **Product facts and prices come from the OKF bundle** —
`references/scalar/okf-bundle.md` tells you where it is and how to cite it.
Capability, edition, version, release status, billing model, list price, and
Pod counting are looked up there, never written from memory. The bundle is a
public repository, so its figures are citable — label them 定価 (tax-excluded)
and treat them as reference-estimate material, with the AE reviewing any quote
that goes to the customer. What it marks 非公開 (3-year terms, prepaid credits,
discounts) routes to the AE instead.
- **Research freshness**: facts come from `references/scalar/research-2026-08.md`
and `references/scalar/proposal-map.md` (§3/§5 dated 2026-08-05). Both follow
the **3-month rule** — refresh only affected claims using
`references/scalar/research-policy.md`; parallel research is not automatic.
## Quick Reference
| Task | Use |
|------|-----|
| Hearing checklist / proposal structure / challenge→product map / constraints / pricing / standard environments + BOM | `references/scalar/proposal-map.md` |
| Hearing forms to hand the AE (fill-in sheets) | `templates/sales/hearing-sheet.ja.md` (product-neutral) + `templates/sales/products/scalar.ja.md` (Scalar fit) |
| A hearing sheet in Excel / Google Spreadsheet, and reading it back | `hearing-sheet` skill (`scripts/hearing/hearing_sheet.py`) |
| Slides that ask the customer for what is still missing | `hearing-slides` skill |
| Proposal deck builder (worked example, 23 slides incl. architecture + BOM) | `scripts/scalar/build_scalar_proposal.py` |
| Environment diagram source (3 environments, AWS) | `examples/scalar-proposal-envs.drawio` → PNG via `scripts/drawio_export.py` |
| Product capability / edition / version / list price / Pod counting | `references/scalar/okf-bundle.md` → the OKF bundle |
| Researched company/product facts + pitfalls | `references/scalar/research-2026-08.md` |
| Section ordering rationale (problem-solving outline) | `references/deck-outlines.md` |
| Run | `cd /path/to/slide-forge && .venv/bin/python scripts/scalar/build_scalar_proposal.py [--folder <Drive URL>]` |
| Validate first (no API calls) | same command with `--dry-run` — runs the coordinate and text-fit audits without creating a deck |
## Phase 1: Collect the challenges and settle premises
Follow `references/interactive-intake.md` sections 0/3/4/5. Ask in one batch:
| # | header | Question | Options |
|---|---|---|---|
| 1 | Challenge material | Do you have material on the customer's challenges? | Provide minutes/hearing notes (file/paste) / Explain verbally now / Not yet (start from presenting the hearing checklist) |
| 2 | Challenge category | Which category are the main challenges closest to? | 3 likely categories from proposal-map.md §3's A–H based on the material + Other (multiple selection allowed) |
| 3 | Proposal stage | What stage is this proposal at? | Initial proposal (mainly to agree on challenges) / PoC proposal (through scope and success criteria) / Full-rollout proposal (cost and team at a fixed level of detail) |
| 4 | Decision maker | Who is the main reader? | Executive / decision maker / IT department / Business department / Mixed |
- Material first: if minutes/notes exist, **read them before asking Q2** and
pre-select the likely categories in the option descriptions.
- If nothing is known yet, hand over the hearing forms as the deliverable
instead of forcing a deck — `templates/sales/hearing-sheet.ja.md` plus
`templates/sales/products/scalar.ja.md` (what each section feeds in the deck
is in the sheet's §14.3; the checklist behind them is proposal-map.md §1).
- If a filled-in hearing sheet exists, read it first: §4.2 / §5 give the
current-state diagram and the BOM quantities, and the product addendum's
§B / §C give the constraints that must appear on the risk slide. When it is
kept as `hearing.json`, run `hearing_sheet.py gaps` before designing — what
is still `未確認` becomes "to confirm today" on the deck, never a guess, and
what is `推定` is stated as our reading rather than as fact.
- Second round if unspecified: output Drive folder, cover date, language, and
cloud (default AWS — state the adopted default instead of asking when the
customer's cloud is unknown). The shared contract owns the QA question.
- **Do not ask about** diagram composition, coordinates, colors, or which
part draws which section — that is fixed by the worked example and
design conventions.
Then **present the slide outline (page count + each slide's action title,
challenge→product mapping made explicit) and get approval before generating**
(the outline gate of interactive-intake.md §3).
## Phase 2: Map challenges to products
1. Classify each agreed challenge into proposal-map.md §3 categories (A–H);
pull the product/feature line and the public cases from the same row.
2. Check §4 (unsuitable cases) — direct-write bypass, DB-specific features,
OLAP-only workloads, "tamper prevention" wording, Community-edition gaps.
Surface any hit as a risk-slide item or descope it honestly.
3. Verify freshness (3-month rule) of §3/§5 and research-2026-08.md; re-run
research agents if stale, and check the pitfall list at the end of
research-2026-08.md before writing slides.
## Phase 2.5: Compose the architecture and BOM
1. Start from the standard 3-environment topology (proposal-map.md §6):
development (local, Community edition, free) → test aidd-infra-test
(Cluster, 1 Pod) → staging aidd-infra-staging (Cluster, 3 Pods), AWS by
default. Adjust environment names, sizes, and the production environment
to the hearing.
2. Author the diagram from `examples/scalar-proposal-envs.drawio` (edit →
`scripts/drawio_export.py` → Read the PNG; drawio-diagrams skill rules
apply: verified shape names only, visual check mandatory).
3. Derive the BOM: per environment, the cloud services and the Scalar
products with quantities; compute monthly license cost with the §6 formula
when quantities are not customer-specified. Premium features or ScalarDL
change the unit prices (§5/§6).
4. **When the customer needs an editable estimate** (itemized quotation) —
typical for full-rollout proposals, or when the user asks — build it from the BOM with
the `spreadsheets` skill (Excel / Google Spreadsheet, real formulas for
quantities × unit prices and tax) into the deck's Drive folder. The cost
slide keeps the summary; the spreadsheet carries the line items, and the
two totals must match.
Section pages that used to exist only inside the builder now have registered
templates — reach for these instead of re-drawing figures by hand:
| Section | Template (pack) |
|---|---|
| Challenge structure | `iceberg-challenge` (`proposal`) |
| Challenge → solution mapping | `challenge-solution-map` (`proposal`) |
| Target state and scope | `scope-in-out` (`proposal`) |
| Expected outcomes | `outcome-before-after` (`proposal`) |
| PoC plan and pass criteria | `poc-plan` (`proposal`) |
| Case studies | `case-study-card` / `case-study-detail` / `case-fit` (`case-studies`) |
| Next steps | `next-step-customer` (`proposal`) |
A case page needs the customer's publication consent and a dated source —
`templates/marketing/case-study.ja.md` §1 and §3 own both. **Do not put a case
on a customer page while consent is 未取得.**
## Phase 3: Build the deck
`scripts/scalar/build_scalar_proposal.py` is a worked example (fictional
manufacturing scenario, 23 slides incl. architecture + BOM) that encodes the
standard structure —
proposal-map.md §2 has the section-by-section rationale. To build a real
proposal, rewrite only the `PROPOSAL` dict at the top of the script
(customer, summary, challenges, mapping, alternatives, effects, cases,
journey, gantt, team, costs, risks, next) with hearing results, keeping:
- Challenge slides ≤ 3 items, same order and wording as the mapping table
- The alternatives table's comparison axes rewritten to the customer's actual
KBF (what they will evaluate proposals on)
- Cases picked from the mapped categories (§3), with sources in speaker notes
- Cost figures only from §5 with `source_note`; anything else stays
"quote available on request"
- Dense architecture diagrams (10+ nodes, cloud-specific) → author with the
`drawio-diagrams` skill and insert as an image instead of the built-in
solution diagram
Design conventions are shared with scalar-product-slides: action titles,
square corners on accent-bar cards, pictograms from `illustrations`.
## Phase 4: Generate and QA
```bash
cd /path/to/slide-forge
.venv/bin/python scripts/scalar/build_scalar_proposal.py --dry-run # audits only, no API
.venv/bin/python scripts/scalar/build_scalar_proposal.py [--folder <URL>]
```
1. The script prints "audit:" lines from `audit_*` on every drawn slide.
**If any audit fires, fix the data/spec and rebuild** (delete the old deck
from Drive first; rebuilding changes the URL — tell the user).
2. Apply the shared contract's QA and cleanup procedure.
3. Content QA specific to proposals: no customer-specific number without a
hearing basis, no case/price without a source note, challenge wording
consistent across slides 3 / 7 (challenge summary and mapping table), scope-out line present.
4. In the final report, alongside the deck/folder URLs, include the BOM
lists (cloud services per environment; Scalar products with quantity and
monthly cost) — the builder prints them as `=== Bill of Materials (BOM) ===`.
5. Use the shared contract's final-adjustment step, adding proposal-specific
choices for swapping cases or adding/removing sections.
## File layout
| Path | Role |
|------|------|
| `scripts/scalar/build_scalar_proposal.py` | Proposal deck builder (worked example; rewrite `PROPOSAL` per customer) |
| `references/scalar/proposal-map.md` | Hearing items, proposal structure + rationale, challenge→product map, constraints, pricing |
| `references/scalar/okf-bundle.md` | Where the OKF bundle is and how to cite product facts and prices from it |
| `references/scalar/research-2026-08.md` | Company/product facts, cases, slide pitfalls (shared with scalar-product-slides) |
| `examples/scalar-proposal-envs.drawio` / `.png` | 3-environment architecture diagram source and export (rewrite per customer) |
| `templates/scalar-2026.json` | Scalar 2026 template |
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!