| title | SDStack |
|---|---|
| description | The open enterprise skill stack — business and IT skills organized by department, company size, and process. For humans and AI agents alike. |
| audience | both |
| category | index |
███████╗██████╗ ███████╗████████╗ █████╗ ██████╗██╗ ██╗ ██╔════╝██╔══██╗██╔════╝╚══██╔══╝██╔══██╗██╔════╝██║ ██╔╝ ███████╗██║ ██║███████╗ ██║ ███████║██║ █████╔╝ ╚════██║██║ ██║╚════██║ ██║ ██╔══██║██║ ██╔═██╗ ███████║██████╔╝███████║ ██║ ██║ ██║╚██████╗██║ ██╗ ╚══════╝╚═════╝ ╚══════╝ ╚═╝ ╚═╝ ╚═╝ ╚═════╝╚═╝ ╚═╝
Every department. Every company size. Every process.
For humans and AI agents alike.
SDStack is a library of skills that an AI agent (or a human) can pick from. It has three top-level concepts:
departments/— work scoped to a specific business function. Pinned to department × company size × skill (e.g.,finance-accounting/02-org-100-to-1k/accounts-payable/SKILL.md).templates/— cross-department work artifacts (e.g., fit-gap matrix, go-live checklist). Scoped by company size only.investment-research/— equity research, market commentary, and stock recommendations. For agents that combine domain expertise with sector knowledge + maintained ticker watchlists to publish investment content. Scoped by sector, not department or size.
Pick the right top-level concept, then the right leaf — you land on exactly one SKILL.md file. No ambiguity, no cross-referencing multiple folders.
skills/
├── departments/ every department-owned skill lives here
│ ├── customer-support/
│ │ ├── OVERVIEW.md department-level context (read first)
│ │ ├── 01-org-under-100/<skill>/SKILL.md
│ │ ├── 02-org-100-to-1k/<skill>/SKILL.md
│ │ └── 03-org-1k-plus/<skill>/SKILL.md
│ ├── finance-accounting/
│ ├── human-resources/
│ ├── information-technology/
│ │ ├── OVERVIEW.md
│ │ ├── role-overviews/ reference: who does what in IT
│ │ │ ├── compliance-analyst.md
│ │ │ ├── data-engineer.md
│ │ │ ├── migration-architect.md
│ │ │ ├── qa-lead.md
│ │ │ ├── requirements-analyst.md
│ │ │ └── solution-architect.md
│ │ ├── 01-org-under-100/
│ │ ├── 02-org-100-to-1k/
│ │ └── 03-org-1k-plus/<skill>/SKILL.md
│ ├── project-operations/
│ ├── sales-crm/
│ └── supply-chain/
│
├── templates/ cross-department skills, size-scoped only
│ ├── 01-org-under-100/<template>/SKILL.md
│ ├── 02-org-100-to-1k/<template>/SKILL.md
│ └── 03-org-1k-plus/<template>/SKILL.md
│
└── investment-research/ equity research + market commentary (sector-scoped)
├── OVERVIEW.md
├── core/ sector-agnostic research craft
│ ├── equity-research-framework/SKILL.md
│ ├── reading-10k-10q/SKILL.md
│ ├── earnings-call-analysis/SKILL.md
│ ├── valuation-dcf-comps/SKILL.md
│ ├── competitive-landscape/SKILL.md
│ └── content-publishing/SKILL.md
└── sectors/ one folder per sector
├── hr-tech/ SKILL.md + tickers.md
├── finance-accounting-tech/
├── sales-crm-tech/
├── customer-support-tech/
├── supply-chain-tech/
├── project-ops-tech/
└── platform-it/
The leaf paths are always:
departments/<department>/<org-size>/<skill>/SKILL.md
templates/<org-size>/<template>/SKILL.md
investment-research/core/<skill>/SKILL.md
investment-research/sectors/<sector>/SKILL.md
investment-research/sectors/<sector>/tickers.md
The size tier refers to the whole organization's headcount, not the size of the department. A 5,000-person company with a 10-person HR team is still 03-org-1k-plus — the company size is what drives payroll complexity, compliance load, HRIS scale, etc.
| Folder | Meaning |
|---|---|
01-org-under-100 |
Fewer than 100 people in the whole org |
02-org-100-to-1k |
100 to 1,000 people in the whole org |
03-org-1k-plus |
1,000+ people in the whole org |
Why the numeric prefix? So folders sort by size, and new tiers can be inserted later (e.g. 04-org-10k-plus/) without renaming any existing folder. If a middle tier ever splits, slot in 02a-org-100-to-500/ and 02b-org-500-to-1k/ alongside — existing references don't break.
| Department | Top-level content |
|---|---|
| Customer Support | Ticketing, SLA, escalation, knowledge base, omnichannel, field service, CSAT |
| Finance & Accounting | GL, AP/AR, fixed assets, budgeting, period close, consolidation, tax |
| Human Resources | Payroll, benefits, recruitment, onboarding/offboarding, leave, performance, training |
| Information Technology | API design, integrations, data modeling/migration, security, observability, DR, testing, and more |
| Project Operations | Project planning, resourcing, time & expense, billing, portfolios, subcontractors |
| Sales & CRM | Leads, pipeline, quoting/CPQ, contracts/renewals, commissions, territory, customer 360, campaigns |
| Supply Chain | Procurement, inventory, warehouse, demand planning, logistics, quality, vendor management, production |
Each department's OVERVIEW.md holds the cross-skill reference (entities, regulatory load, integration shape, KPIs). Read it before diving into any individual skill under that department.
Inside information-technology/, the role-overviews/ folder holds reference docs for the IT implementation roles (solution architect, data engineer, migration architect, QA lead, requirements analyst, compliance analyst). These are not skills themselves — they're how to think like that role when executing IT skills.
Cross-department artifacts. Scoped by org size only — the same artifact looks very different at 50 people vs. 5,000.
templates/03-org-1k-plus/fit-gap-matrix/templates/03-org-1k-plus/go-live-checklist/templates/03-org-1k-plus/migration-runbook/templates/03-org-1k-plus/requirements-traceability/
Smaller-org versions are scaffolded empty; fill them as the need arises.
A top-level concept for agents producing equity research, market commentary, and stock recommendations. Combines sector-agnostic research craft (reading filings, valuation, thesis-building, publishing) with sector-specific knowledge + maintained ticker watchlists via Google Finance.
Shape:
investment-research/
├── OVERVIEW.md
├── core/ 6 research-craft skills
└── sectors/ 7 sectors with SKILL.md + tickers.md each
├── hr-tech/ ~60 tickers (HCM, talent, staffing, L&D, benefits)
├── finance-accounting-tech/ ~90 tickers (ERP, AP, close, FP&A, tax, payments, banking)
├── sales-crm-tech/ ~55 tickers (CRM, marketing, ad-tech, CDPs)
├── customer-support-tech/ ~25 tickers (helpdesk, CCaaS, VoC, field service)
├── supply-chain-tech/ ~70 tickers (WMS, freight, trucking, rail, parcel)
├── project-ops-tech/ ~30 tickers (work mgmt, IT services, consulting)
└── platform-it/ ~110 tickers (cloud, security, observability, data, AI)
An agent composes: domain expertise (from departments/) + research craft (from core/) + sector knowledge + ticker watchlist (from sectors/<sector>/). Example: an HR-expert agent picks departments/human-resources/ + investment-research/core/* + investment-research/sectors/hr-tech/ and publishes commentary on Workday earnings, Paycom vs Paylocity, or the Rippling IPO. See investment-research/OVERVIEW.md for the composition pattern.
Public-ticker universe: ~400+ across the 7 sectors, covering the sizable majority of public enterprise-tech + adjacent ecosystem.
- Who are you helping? Identify department (or "this is cross-department" → templates).
- How big is their org? Pick
01-org-under-100,02-org-100-to-1k, or03-org-1k-plus. An agent scoped to a size stays at that size — sizes are a trait, not a trajectory. - What's the specific skill? Navigate to the leaf
SKILL.md.
The whole skill — context, decisions, build steps, integrations, checklists — lives in that one file. No need to assemble from multiple sources.
To improve a skill, edit its SKILL.md in place. To add a new skill, create departments/<dept>/<size>/<new-skill>/SKILL.md following the frontmatter format used by existing files.
- Business user: Find your department → your org size → your process. The
SKILL.mdexplains what to do, how to set it up, what to watch. - Builder / architect: Start at the department
OVERVIEW.mdto understand the terrain, then drop into individual skills. - AI agent operator: Treat each
SKILL.mdas a self-contained mission context. Frontmatter is machine-parseable.
Today, every department has its 03-org-1k-plus/ populated (the enterprise tier is the deepest). 01-org-under-100/ and 02-org-100-to-1k/ folders exist as empty scaffolds and will fill in over time. Same story for templates/.
MIT — use it, fork it, build with it.
SD — Shashank Dixit · ERP•AI