Ripple

Release impact analyzer for QA/SDET engineers. Given a Jira ticket or release, Ripple cross-references your Confluence documentation and existing test suite to produce a structured report: what changed, what's at risk, which tests to run, and where the coverage gaps are.

Installation

npm install -g ripple-qa

Quick Start

# 1. Run the setup wizard
ripple init

# 2. Add your API keys to .env
cp .env.example .env
# edit .env with your keys

# 3. Analyze a ticket
ripple analyze --ticket PROJ-1234

Example Output

# Ripple Analysis — PROJ-1234

Add rate limiting to the /payments API endpoint

Analyzed: 2024-03-15T14:32:00.000Z

## Risk Level: HIGH

This is a high-priority Bug affecting the payments component, which is core to checkout flows.

## Impacted Areas

| Area | Confidence | Reason |
|------|-----------|--------|
| Payments API | HIGH | Directly modified endpoint |
| Checkout Flow | HIGH | Payments API is called during checkout |
| Fraud Detection | MEDIUM | Rate limiting may interact with retry logic in fraud checks |

## Recommended Tests (6)

### HIGH Priority
- **TC-001: Successful payment with valid card** (Payments)
  → Core happy path for the modified endpoint
- **TC-002: Payment declined — insufficient funds** (Payments)
  → Ensures error handling still functions under rate limiting

### MEDIUM Priority
- **TC-015: Checkout end-to-end flow** (Checkout)
  → Rate limiting could break checkout if limits are too aggressive

## Coverage Gaps

These scenarios have no existing test case:

1. **Rate limit threshold is enforced (e.g. >100 req/min returns 429)**
   Suggested: Verify /payments returns HTTP 429 after exceeding rate limit within a time window

2. **Rate limit resets correctly after the window expires**
   Suggested: Confirm a client blocked by rate limiting can retry successfully after the cooldown period

## Context

- Wiki pages referenced: Payments Architecture, Checkout Flow Overview
- Test cases evaluated: 87
- Test cases recommended: 6

---
Generated by Ripple • claude-sonnet-4-6

CLI Reference

# Single ticket
ripple analyze --ticket PROJ-1234

# Multiple tickets
ripple analyze --ticket PROJ-1234 PROJ-1235 PROJ-1236

# Full release
ripple analyze --release v2.4.1

# Override output format
ripple analyze --ticket PROJ-1234 --output json
ripple analyze --ticket PROJ-1234 --output both

# Save report to file
ripple analyze --ticket PROJ-1234 --save
# Saves to ./ripple-reports/PROJ-1234-<timestamp>.md

# Verbose mode (shows raw source data)
ripple analyze --ticket PROJ-1234 --verbose

Configuration Reference

All non-secret configuration lives in ripple.config.json (safe to commit).

| Field | Description | |-------|-------------| | jira.url | Your Atlassian Cloud base URL (e.g. https://yourcompany.atlassian.net) | | jira.email | Email address associated with your Jira API token | | jira.projectKey | Jira project key (e.g. PROJ) | | confluence.url | Confluence base URL (usually same as Jira for Atlassian Cloud) | | confluence.spaceKey | Confluence space key to search for related docs | | testSuite.type | Currently only csv is supported | | testSuite.path | Relative path to your CSV file from the project root | | testSuite.columns.name | CSV column header for the test case name | | testSuite.columns.area | CSV column header for the feature area | | testSuite.columns.priority | CSV column header for the priority (High/Medium/Low) | | llm.provider | LLM provider: claude (default), github, openai, ollama | | llm.model | Model ID — default depends on provider (e.g. claude-sonnet-4-6, llama3.1:8b) | | llm.baseURL | API base URL — required for ollama, optional for openai | | llm.apiKeyEnv | Env variable name holding the API key — not used for ollama | | output.format | Default output format: markdown, json, or both | | output.saveReports | If true, always saves reports without needing --save | | output.reportsDir | Directory to write saved reports — default: ./ripple-reports |

Environment Variables

Store API keys in a .env file (never commit this file).

| Variable | Description | |----------|-------------| | ANTHROPIC_API_KEY | API key from console.anthropic.com | | JIRA_API_TOKEN | Jira API token from your Atlassian account settings | | CONFLUENCE_API_TOKEN | Confluence API token (usually the same as your Jira token for Atlassian Cloud) |

Supported Sources

| Source | v1 | Planned | |--------|----|---------| | Jira Cloud | ✅ | | | Jira Server/Data Center | | ✅ | | Confluence Cloud | ✅ | | | Confluence Server | | ✅ | | CSV test suite | ✅ | | | TestRail | | ✅ | | Zephyr Scale | | ✅ | | GitHub Issues | | ✅ |

Supported LLM Providers

| Provider | Notes | |----------|-------| | Claude (Anthropic) | Default — set ANTHROPIC_API_KEY | | GitHub Models | Uses your GitHub token — set GITHUB_TOKEN | | OpenAI / compatible | Set OPENAI_API_KEY (or custom env) | | Ollama | Local, no API key — see Using Ollama |

Using Ollama (Local Models)

Ripple supports Ollama as a fully local, no-cost LLM option — no API key required.

1. Install and start Ollama

# macOS / Linux
brew install ollama   # or download from ollama.com
ollama serve

2. Pull a model

ollama pull llama3.1:8b      # recommended — strong reasoning, fast
ollama pull mistral           # good alternative
ollama pull gemma2:9b         # Google's Gemma 2

3. Run ripple init and choose Ollama

? LLM provider: Ollama (local, no API key required)
? Ollama model: llama3.1:8b
? Ollama API base URL: http://127.0.0.1:11434/v1

This writes the following to ripple.config.json:

{
  "llm": {
    "provider": "ollama",
    "model": "llama3.1:8b",
    "baseURL": "http://127.0.0.1:11434/v1"
  }
}

4. Add only Jira/Confluence tokens to .env — no LLM key needed:

JIRA_API_TOKEN=...
CONFLUENCE_API_TOKEN=...

5. Analyze as usual

ripple analyze --ticket PROJ-1234

Note: Local model quality varies. llama3.1:8b works well for structured JSON output. If you get JSON parse errors, try a larger model (llama3.1:70b) or set "model": "mistral" in ripple.config.json.

Contributing

1. Fork the repository
2. Create a feature branch: git checkout -b feat/your-feature
3. Make your changes — ES modules throughout, no TypeScript, no require()
4. Open a pull request

License

MIT