datarails-financeos
Datarails FinanceOS plugin for financial planning and analysis - 17 skills, 8 commands, 8 agents with remote MCP connector for Datarails pla
Open source Repository Open in the app JSON README (API)
About
Datarails FinanceOS plugin for financial planning and analysis - 17 skills, 8 commands, 8 agents with remote MCP connector for Datarails platform.
Details
- Kind
- Plugins
- Topic
- No topic detected
- Publisher
- datarails
- Origin
- marketplace
- Category
- ferramentas
- Stars
- 3
- Forks
- 3
- Last push
- 2026-08-30T06:56:31Z
- Repository state
- ativo
- License
- MIT
- Added
- 2026-08-30 01:48:58
- Updated
- 2026-08-30 01:48:58
- Origin id
datarails/dr-claude-code-plugins-re/datarails-financeos
README
# Datarails Finance OS Plugin for Claude Analyze financial data, detect anomalies, and build FP&A reports directly in Claude — connected live to your Datarails Finance OS. ## Install ### Claude Desktop (Cowork) 1. Open Claude Desktop → **Browse plugins** → **Personal** tab 2. Click **+** → **Add marketplace from GitHub** 3. Enter `Datarails/dr-claude-code-plugins-re` 4. Install **Datarails FinanceOS** > **If marketplace install fails** (a [known SSH issue](https://github.com/anthropics/claude-code/issues/26588)): download the ZIP from the [latest release](https://github.com/Datarails/dr-claude-code-plugins-re/releases/latest) and add it via **+** → **Upload plugin**. ### Claude Code (Terminal) ``` /plugin marketplace add https://github.com/Datarails/dr-claude-code-plugins-re.git /plugin install datarails-financeos@datarails-marketplace ``` > Use the full HTTPS URL (not the `owner/repo` shorthand) — the shorthand triggers SSH cloning, which fails without SSH keys configured. Manage it anytime: `/plugin` (view, enable/disable), `/plugin update`, `/plugin uninstall`. For local development, load a clone directly with `claude --plugin-dir ./dr-claude-code-plugins-re` (that session only). ## Connect to Datarails The Datarails connector is **bundled with the plugin** — there's no separate setup or `claude mcp add` step. The first time a skill needs data, a browser window opens for **OAuth login**; sign in once and you're connected. (In Claude Desktop you can also connect from **+** → **Connectors** → **Datarails**.) ## Using the plugin **Just ask.** Describe what you want in plain language and Claude picks the right skill automatically: > "Give me a financial summary for 2026." > "What are our top expenses this year — flag anything unusual." > "Build an FP&A intelligence workbook." **Or invoke a skill explicitly.** In Claude Code, type `/datarails-financeos:` and press **Tab** to autocomplete every command: ``` /datarails-financeos:tables # list & explore tables /datarails-financeos:financial-summary # revenue / expense / margin snapshot /datarails-financeos:intelligence --year 2026 # full FP&A workbook ``` > The `datarails-financeos:` prefix is **required** — Claude Code namespaces every plugin skill to prevent collisions, so a bare `/tables` won't resolve. ## Skills Invoke any of these as **`/datarails-financeos:<name>`**, or just describe the task and let Claude choose. Every skill discovers your org's tables, fields, scenarios, and account structure at runtime — nothing is hardcoded. ### Explore & query | Skill | What it does | |-------|--------------| | `tables` | List tables, view schemas, inspect a field's values | | `profile` | Field statistics — null rates, cardinality, range outliers | | `query` | Fetch and filter records; value-list **and** advanced filters (ranges, comparisons, text, null checks, date ranges) | | `test` | Check which fields work as aggregation dimensions in your environment | ### Snapshots & trends | Skill | What it does | |-------|--------------| | `financial-summary` | Period-scoped snapshot: revenue, expenses, gross profit & margin | | `revenue-trends` | Revenue growth rates, seasonality, composition over time | | `expense-analysis` | Top expense categories, concentration, monthly trend | ### Reports & workbooks | Skill | What it does | Output | |-------|--------------|--------| | `intelligence` | Full-fiscal-year FP&A workbook with auto-detected insights (no PowerPoint) | Excel (up to 10 sheets) | | `insights` | Executive trend analysis and narrative | PowerPoint + Excel | | `dashboard` | Executive KPI dashboard | Excel + PowerPoint | | `departments` | Departmental P&L and comparison | Excel + PowerPoint | | `forecast-variance` | Multi-scenario variance — scenarios discovered at runtime (no fixed "Budget" assumed) | Excel + PowerPoint | | `extract` | Extract validated financial data (P&L, Balance Sheet, sourceable KPIs) | Excel | ### Assurance & data quality | Skill | What it does | Output | |-------|--------------|--------| | `anomalies` | Detect data-quality anomalies in a table | inline | | `anomalies-report` | Full data-quality assessment with a quality score | Excel | | `reconciliation` | Independent-source consistency checks — cross-endpoint agreement, balance-sheet identity, roll-ups, scenario/period integrity. Validates the **data pipeline**, not source systems | Excel | | `audit` | Audit-support evidence package (completeness, reconciliation, mapping integrity, sampling). **Not** a SOX certification — access-control / change-management / ITGC evidence is out of scope | PDF + Excel | ### Excel & drill-down | Skill | What it does | |-------|--------------| | `get-formula` | Generate Excel workbooks with live `DR.GET` formulas | | `drilldown` | Break a Datarails number into its line items — from a workbook cell, a pasted `DR.GET` formula, or a plain-language description (no file needed) | > A few short **command aliases** also resolve under the same namespace for common tasks — `/datarails-financeos:explore-tables`, `:data-check`, `:budget-comparison`, `:test-api` — each launches the matching skill above. ## The intelligence workbook The flagship report. Run it for a full FP&A book: ``` /datarails-financeos:intelligence --year 2026 ``` It generates **up to 10 sheets** — conditional sheets are included only when your data sources them, and omitted (never fabricated) otherwise: 1. **Insights Dashboard** — top findings with severity 2. **Expense Deep Dive** — top accounts, % of total 3. **Variance Waterfall** — what changed and why 4. **Trend Analysis** — 12-month trends 5. **Anomaly Report** — auto-detected outliers 6. **Vendor Analysis** — top vendors, concentration *(when vendor data exists)* 7. **SaaS Metrics** — ARR, NRR, CAC, LTV *(when your data sources them)* 8. **Sales Performance** — rep leaderboard *(when rep/bookings data exists)* 9. **Cost Center P&L** — department detail 10. **Raw Data** — pivot-ready ## Authentication Handled automatically via **OAuth 2.0 + PKCE** when the connector is first used — a browser window opens for login, tokens refresh on their own, and there are no manual steps. ## Troubleshooting | Problem | Fix | |---------|-----| | A command isn't found | Type `/datarails-financeos:` and press **Tab** — every command is under that prefix; a bare name won't resolve | | Tools not available | Connect the Datarails connector: **+** → **Connectors** → **Datarails** → **Connect** | | "Not authenticated" | Disconnect and reconnect the Datarails connector | | Skill picked the wrong table/field | Skills discover your financials table and fields automatically; if it guesses wrong, tell it which to use and it continues — there's no profile to configure | | Commands not showing | Restart Claude Desktop / Claude Code, and check the plugin is enabled (`/plugin`) | | Slow extraction | Row-by-row extraction pages 500 rows at a time and can take minutes on large tables; summary skills use aggregation, which returns in seconds | See [SETUP.md](SETUP.md) for detailed setup and troubleshooting, or the [Getting Started Guide](docs/guides/GETTING_STARTED.md) for a ~15-minute hands-on walkthrough. ## License MIT License — see LICENSE file. ## Privacy This plugin connects to the Datarails Finance OS service to analyze your financial data. See the [Datarails Privacy Policy](https://www.datarails.com/privacy-policy/) for how your data is collected, used, and protected. ## Support - GitHub Issues: https://github.com/Datarails/dr-claude-code-plugins-re/issues - Datarails Support: support@datarails.com