Initial release: AI-native Robinhood trading interface

TypeScript monorepo with two packages:
- @rh-agent-tools/client: Standalone Robinhood API client (~50 async methods)
- rh-agent-tools: MCP server with 18 structured tools

Features:
- Browser-based authentication via Playwright (Chrome)
- AES-256-GCM encrypted session storage with OS keychain
- Multi-account support across all account-scoped methods
- 5 Claude Code skills (setup, portfolio, research, trade, options)
- Safety controls: blocked fund transfers, blocked bulk cancels, explicit order params
- Compatible with Claude Code, Codex, OpenClaw, and any MCP-compatible agent

Author: kevin1chun

.editorconfig

@@ -0,0 +1,11 @@
+root = true
+
+[*]
+end_of_line = lf
+insert_final_newline = true
+charset = utf-8
+trim_trailing_whitespace = true
+
+[*.{ts,js,json,md,yml,yaml}]
+indent_style = space
+indent_size = 2

.github/CODEOWNERS

@@ -0,0 +1 @@
+* @kevin1chun

.github/ISSUE_TEMPLATE/bug_report.md

@@ -0,0 +1,34 @@
+---
+name: Bug Report
+about: Report a bug or unexpected behavior
+labels: bug
+---
+
+## Environment
+
+- **rh-agent-tools version**:
+- **Bun version** (`bun --version`):
+- **OS**:
+- **MCP client** (Claude Code / Claude Desktop / other):
+
+## Description
+
+A clear description of the bug.
+
+## Steps to Reproduce
+
+1.
+2.
+3.
+
+## Expected Behavior
+
+What you expected to happen.
+
+## Actual Behavior
+
+What actually happened. Include error messages or MCP tool output if applicable.
+
+## Security Notice
+
+**Never paste tokens, credentials, session data, or account numbers in issues.**

.github/ISSUE_TEMPLATE/feature_request.md

@@ -0,0 +1,21 @@
+---
+name: Feature Request
+about: Suggest a new tool, skill, or improvement
+labels: enhancement
+---
+
+## Use Case
+
+Describe the problem or workflow this feature would support.
+
+## Proposed Solution
+
+What you'd like to see (new MCP tool, client method, skill, etc.).
+
+## Safety Considerations
+
+Does this feature involve order placement, account modifications, or sensitive data? If so, describe the safety implications and any safeguards you'd suggest.
+
+## Alternatives Considered
+
+Any alternative approaches you've considered.

.github/pull_request_template.md

@@ -0,0 +1,17 @@
+## Summary
+
+Brief description of the changes.
+
+## Safety Checklist
+
+- [ ] Does not expose fund transfer or bank operations
+- [ ] Does not enable bulk operations without safeguards
+- [ ] Order-related changes require explicit parameters (no dangerous defaults)
+- [ ] ACCESS_CONTROLS.md updated if adding new operations
+
+## Testing
+
+- [ ] `npx vitest run` passes
+- [ ] `bun run typecheck` passes
+- [ ] `bun run check` passes
+- [ ] New tests added for new functionality

.github/workflows/ci.yml

@@ -0,0 +1,18 @@
+name: CI
+
+on:
+  push:
+    branches: [master]
+  pull_request:
+    branches: [master]
+
+jobs:
+  check:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: oven-sh/setup-bun@v2
+      - run: bun install --frozen-lockfile
+      - run: bun run check
+      - run: bun run typecheck
+      - run: npx vitest run

.gitignore

@@ -0,0 +1,9 @@
+node_modules/
+dist/
+.env
+*.enc
+.vscode/
+.idea/
+.DS_Store
+*~
+coverage/

CHANGELOG.md

@@ -0,0 +1,23 @@
+# Changelog
+
+All notable changes to this project will be documented in this file.
+
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
+and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## [0.3.0] - 2025-03-10
+
+### Added
+
+- **MCP Server** with 18 structured tools for any MCP-compatible AI agent
+- **Standalone client library** (`@rh-agent-tools/client`) with ~50 async methods
+- **5 Claude Code skills**: setup, portfolio, research, trade, options
+- Browser-based authentication via Playwright (Chrome)
+- AES-256-GCM encrypted session storage with OS keychain key management
+- Multi-account support (first-class across all account-scoped methods)
+- Interactive onboarding TUI (`rh-agent-tools onboard`)
+- One-command install for Claude Code (`rh-agent-tools install`)
+- Safety controls: blocked fund transfers, blocked bulk cancels, explicit order parameters
+- Support for Claude Code, Codex, and OpenClaw agents
+
+[0.3.0]: https://github.com/kevin1chun/rh-agent-tools/releases/tag/v0.3.0

CLAUDE.md

@@ -0,0 +1,79 @@
+# rh-agent-tools
+
+AI-native Robinhood trading interface — TypeScript monorepo with a standalone API client and MCP server.
+
+## Project Structure
+- `packages/client/` — `@rh-agent-tools/client`: Standalone Robinhood API client
+- `packages/server/` — `rh-agent-tools`: MCP server with 18 tools
+- `.claude/skills/` — Claude Code skills for interactive use (SKILL.md only, no scripts)
+- `docs/` — Architecture, access controls, use cases, contributing
+
+## Tech Stack
+- **Runtime**: Bun
+- **Language**: TypeScript (strict mode, ESM-only)
+- **MCP SDK**: `@modelcontextprotocol/sdk` v1.12+ (McpServer + StdioServerTransport)
+- **Validation**: Zod v3.24 (API responses + MCP tool schemas)
+- **Testing**: Vitest (not `bun test` — module isolation matters)
+- **Linting**: Biome v2
+- **Browser Auth**: playwright-core (browser auth)
+
+## Running the MCP Server
+```bash
+bun install
+bun packages/server/bin/rh-agent-tools.ts
+```
+
+## Development
+```bash
+bun run typecheck   # tsc --noEmit on both packages
+bun run check       # biome lint + format
+npx vitest run      # all tests (use vitest, NOT bun test)
+```
+
+## Skills
+Canonical skill source is `packages/server/skills/`. Local `.claude/skills/` contains symlinks for development.
+
+Install MCP server + skills: `bun packages/server/bin/rh-agent-tools.ts install`
+
+Skills use three-layer progressive disclosure:
+1. **SKILL.md** — MCP tool orchestration (default)
+2. **reference.md** — MCP tool API details (loaded on demand)
+3. **client-api.md** — TypeScript client library patterns (advanced, loaded on demand)
+
+Available skills:
+- `robinhood-setup` - Initial auth setup
+- `robinhood-portfolio` - Portfolio and holdings
+- `robinhood-research` - Stock research and analysis
+- `robinhood-trade` - Order placement with safety checks
+- `robinhood-options` - Options chain analysis
+
+## Client Patterns
+```typescript
+import { RobinhoodClient, getClient } from "@rh-agent-tools/client";
+
+// Class-based
+const client = new RobinhoodClient();
+await client.restoreSession();
+const quotes = await client.getQuotes("AAPL");
+
+// Singleton
+const rh = getClient();
+await rh.restoreSession();
+```
+- All methods are `async` (native `fetch` under the hood)
+- Multi-account is first-class: every account-scoped method accepts `accountNumber`
+- Session cached to `~/.rh-agent-tools/session.enc` (AES-256-GCM, key in OS keychain)
+- Proper exceptions: `AuthenticationError`, `APIError`
+- **Do NOT use `phoenix.robinhood.com`** — it rejects TLS. Use `api.robinhood.com` endpoints only.
+
+## Safety Rules
+- **NEVER** place bulk cancel operations
+- **NEVER** call fund transfer functions
+- **ALWAYS** confirm with user before placing any order
+- Order tools require explicit parameters - no defaults that could cause accidental trades
+
+## Testing
+```bash
+npx vitest run
+```
+Tests use `vi.mock()` to mock HTTP layer — no real API calls. Use `vitest` (not `bun test`) for correct module isolation.

CODE_OF_CONDUCT.md

@@ -0,0 +1,47 @@
+# Contributor Covenant Code of Conduct
+
+## Our Pledge
+
+We as members, contributors, and leaders pledge to make participation in our
+community a welcoming experience for everyone.
+
+## Our Standards
+
+Examples of behavior that contributes to a positive environment:
+
+- Using welcoming and inclusive language
+- Being respectful of differing viewpoints and experiences
+- Gracefully accepting constructive criticism
+- Focusing on what is best for the community
+- Showing empathy towards other community members
+
+Examples of unacceptable behavior:
+
+- Trolling, insulting or derogatory comments, and personal or political attacks
+- Public or private harassment
+- Publishing others' private information without explicit permission
+- Other conduct which could reasonably be considered inappropriate in a
+  professional setting
+
+## Our Responsibilities
+
+Project maintainers are responsible for clarifying the standards of acceptable
+behavior and are expected to take appropriate and fair corrective action in
+response to any instances of unacceptable behavior.
+
+## Scope
+
+This Code of Conduct applies within all project spaces, and also applies when
+an individual is representing the project in public spaces.
+
+## Enforcement
+
+Instances of unacceptable behavior may be reported to the project team at
+**kevin@taji.finance**. All complaints will be reviewed and investigated
+promptly and fairly.
+
+## Attribution
+
+This Code of Conduct is adapted from the [Contributor Covenant](https://www.contributor-covenant.org),
+version 2.1, available at
+https://www.contributor-covenant.org/version/2/1/code_of_conduct.html.