What does this skill do, and when should you use it?
RiseUp is an Israeli personal finance app with no public API. This open-source CLI reverse-engineers its internal interface so you can query spending, income, bank balances, credit card debt, and trends from the terminal, or let AI agents like Claude Code answer finance questions in natural language. All requests go directly from your machine to RiseUp using your own session cookies, with nothing proxied. It is a community project, not affiliated with RiseUp.
Once installed, it runs riseup commands or natural-language agent queries: spending by category/merchant/payment source, monthly income and salary, transaction search and management (classify, rename, comment, exclude from budget, billing merges), bank balances and investment portfolio, credit card debt, multi-month trend comparisons, financial health recommendations, savings plans, AI insights, and bank sync status. Output is a colored table or --; authentication is Google OAuth through a real Chrome driven by Playwright.
- An Israeli RiseUp user who wants to see this month's spending by category from the terminal
- A Claude Code user asking 'how much did I spend this month?' in natural language
- Someone batch-classifying unclassified transactions and teaching auto-matching with `--apply-to all`
- Users analyzing subscriptions (recurring merchants across months) or comparing multi-month income vs. expenses
- Users checking whether their bank connections are syncing and reviewing credit card debt
How do you install this skill?
- The skill relies on reverse-engineered undocumented RiseUp APIs that may break or trigger platform risk controls at any time; account restriction is a third-party risk.
- The OpenClaw install instructions ask users to paste their full session cookie (equivalent to login credentials) into chat — highly sensitive and should never be transmitted via conversation.
- Reachability of core services (RiseUp, Google OAuth, npm, Playwright) from mainland China is unverified; Chinese users may be entirely unable to use this skill.
- Write operations (especially irreversible riseup merge) have confirmation rules, but actual execution behavior is unverified by static review; trial on a small scale first.
- Publisher identity is unverified; this is a community project with no assurance of long-term maintenance or updates.
- Shell / CLI
- Network access
- Local filesystem
Node.js 22+riseup-cli (npm)Playwright ChromiumGoogle Chrome
npm install -g riseup-cli(ornpm install -g github:arsolutioner/riseup-cli); 2.npx playwright install chromium; 3.riseup loginopens Chrome for Google OAuth; 4. verify withriseup status. For Claude Code runriseup skill install; for Claude Desktop downloadriseup-skill.zipfrom GitHub Releases and upload it in Settings > Customize > Skills.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/arsolutioner/riseup-cli.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/src/data" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Once installed, send your agent any of these to trigger it:
- What are my subscriptions?
- Am I saving money?
Example commands: riseup spending, riseup spending --by merchant --top 10, riseup income --salary-only, riseup balance (balances and investments), riseup trends 6, riseup transactions -- for programmatic analysis. With an agent, just ask 'What are my subscriptions?' or 'Am I saving money?'. Note: write operations require confirmation; budget months may not align with calendar months, so verify the actual date range of returned data when querying a specific month.
What are this skill's strengths and limitations?
- Fills the gap left by RiseUp's lack of a public API via reverse engineering
- Data stays local: requests go straight from your machine to RiseUp; session file chmod 0600
- MIT-licensed, TypeScript, dual table/JSON output for scripting
- Tiered command governance: read commands run automatically, write operations require explicit confirmation
- Thorough docs including an error-handling table and the budget-month offset pitfall
- Uses undocumented internal APIs that RiseUp can change at any time, breaking the tool
- Login requires a real Chrome browser (Google OAuth blocks automation), so no fully headless setup
- Data goes back only ~10 months; category names are in Hebrew, requiring translation for English users
- Budget months may not align with calendar months, needing manual date-range checks
- Unofficial and unaffiliated with RiseUp — use at your own risk; the repo shows no test suite
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| RiseUp Finance CLI this page | 52 · Use with care | ★ 26 | 7mo ago | MIT |
| Quicken Mac Read-Only Query Skill | 64 · Recommended | ★ 45 | 1mo ago | MIT |
| Canadian Personal Finance Planner Skill | 58 · Recommended | ★ 63 | 3mo ago | MIT |
| OfficeCLI Financial Modeling Skill | 53 · Use with care | ★ 32k | 12d ago | Apache-2.0 |
| Deep Stock Analysis Workflow (UZI-Skill deep-analysis) | 50 · Use with care | ★ 7.2k | 1mo ago | MIT |
The README names no competing tools; the only alternative for accessing RiseUp data is logging into the web app manually, which is precisely what this CLI eliminates.
How did FollowSkills review this skill?
SKILL.md establishes three autonomy tiers (read-only auto-run, login/logout confirmed, write ops with explicit confirmation and shown commands); README discloses data flow: requests go directly to RiseUp, session stored chmod 0600, no proxying. Deducted for: reliance on reverse-engineered undocumented APIs, unverified publisher identity, and an OpenClaw instruction asking users to paste full session cookies into chat (sensitive credential via conversation, no warning), plus no rollback guidance (merge is irreversible).
Internally consistent docs: command tables, flags, month formats, and an error-handling table (session expiry, timeouts, budget-month offset); JSON mode well defined. Deducted for: no committed test files covering key paths visible in a static read (vitest configured only), no execution evidence, so abnormal-input behavior and failure-feedback quality remain inferred.
Clear triggers (/riseup, finance keywords), disclosed boundaries and known limitations (~10 months data, real Chrome required, budget vs calendar month offset, Hebrew categories). Deducted for: core function depends entirely on the Israeli RiseUp service and Google OAuth — mainland-China reachability unevaluated and likely poor; unusable for non-RiseUp users, with no geographic applicability statement.
Well-layered SKILL.md (prerequisites, autonomy rules, quick reference, workflows, limitations), MIT license, version 0.4.1, npm publish pipeline and docs site. Deducted for: no changelog or explicit update path, maintenance responsibility rests on an unverified individual, and reliance on undocumented APIs limits long-term update assurance.
Commands cover querying, analysis, and writes; JSON mode supports programmatic use; workflows are concrete (subscription detection, multi-month compare, budget-month offset handling), offering marginal value over manual web use. Deducted for: static review cannot verify output correctness or direct usability, no comparison to alternatives, and uncontrollable internal APIs put long-term effectiveness in doubt.
README, SKILL.md, package., and CI workflows cross-corroborate the command set and publish flow, with auditable primary material. Deducted for: no committed test suite or third-party execution evidence in supplied files; CI covers only docs deploy and npm publish, not the skill's key paths; effectiveness claims are author-only.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →