CriticalContext
⚡ 2-minute setup

Your CTO Your product manager Your support team
asks tough questions.
Your code answers. In Slack From an embedded chat From a REST API

Code intelligence for decision makers.
Ask your codebase, not your engineers.

  • Let your PMs, support, execs ask questions about your code directly
  • Free up your valuable development time
  • Powered by Claude Code
Supports: Anthropic API AWS Bedrock Google Vertex AI
Sign Up Free → Or use the prompt below
# eng-questions
PM
Sarah • Product 2:34 PM
@context how do we handle webhook retries?
CX
critical.cx APP 2:34 PM
Webhook retries are handled in app/jobs/webhook_retry_job.rb. The system uses exponential backoff with a max of 5 attempts. Configuration is in config/initializers/webhooks.rb.
Response time: 7.2s
PM
Sarah • Product 2:35 PM
What happens if all 5 attempts fail? Do we log it somewhere?
CX
critical.cx APP 2:35 PM
Yes! After all retries fail, the error is logged in Amplitude as part of the user's journey (event: webhook.delivery.failed). It's also saved to the webhook_failures table in the database.
Response time: 4.8s
📄 Setup: Paste this prompt into Claude Code
Help me set up critical.cx for my team. Before you start, read the API specification from the <script id="context-setup-api"> tag on this page. It contains all the endpoint details and the complete setup flow. Then guide me through each step: 1. Create an account at critical.cx 2. Connect my GitHub repository 3. Set up Slack integration for code queries 4. Start indexing my codebase Open my browser when needed for OAuth flows (GitHub and Slack). Let's start! Ask me for my email and company name to create an account.
~2min setup • Fully transparent • No hidden steps
→
critical.cx setup
$ █
Prefer traditional signup? Create an Account
BUILT FOR ENGINEERS

Enterprise-grade from day one

Your codebase, accessible.

Complete Isolation

Security

Claude Code analyzes your code in our sandboxed worker queue with complete network isolation.

Network isolation

Bring Your Own Keys

Flexibility

We use the Anthropic Claude Code SDK, so API keys, AWS Bedrock, or Google Vertex AI are supported. No pricing markup on LLM usage.

Anthropic Bedrock Vertex AI

Multiple Integrations

Integrations

Slack bot, REST API, embeddable chat widget for your internal websites.

Slack Chat Widget REST API
WHO IT'S FOR

Everyone gets answers.
Engineers stay focused.

The same codebase. Different questions. Instant answers.

EXECUTIVES & LEADERSHIP

CEOs, CTOs, VPs — get the data you need for strategic decisions

Measure Technical Debt for Board Decks

"What percentage of our codebase was written over 3 years ago and hasn't been touched? Which modules have the highest cyclomatic complexity?" Get hard numbers for quarterly planning and investor updates.

Ask: @context How much of our payment processing code is legacy Rails 5 code?
Why it matters: Justify engineering investments with real data, not gut feelings

Validate M&A Technical Due Diligence

"Does this acquisition target have proper test coverage? Are there hardcoded credentials? What's their dependency health?" Get pre-meeting intelligence on codebases you're evaluating.

Ask: @context What security vulnerabilities exist in our authentication flow?
Why it matters: Spot red flags before signing term sheets

Track Engineering Velocity Metrics

"How many new features shipped this quarter vs. bug fixes? Which teams are moving fastest? Where are we spending cycles on rework?" Tie code activity to OKRs and KPIs.

Ask: @context Show me git commit patterns for the checkout team this month
Why it matters: Measure impact without waiting for eng managers to compile reports
PRODUCT MANAGERS

PMs, TPMs, Product Ops — scope features and manage timelines with precision

Scope Feature Complexity Before Sprint Planning

"If I want to add Apple Pay to checkout, what existing code can we reuse? How many files would need to change?" Get real estimates before committing to roadmaps.

Ask: @context What would it take to add Apple Pay? Show me the payment integration points.
Why it matters: Stop underscoping features in planning meetings

Map User Journeys to Actual Code Paths

"When a user clicks 'Buy Now', what happens? Which services fire? Where could this fail?" Trace the full funnel from button click to confirmation email.

Ask: @context Walk me through the entire checkout flow from cart to order confirmation
Why it matters: Debug conversion drops without interrupting engineers

Understand Feature Flag Coverage

"Which features are behind flags? What's the rollout status? How do I enable the new pricing page for beta users?" Self-serve feature management questions.

Ask: @context List all feature flags related to the pricing page and their current states
Why it matters: Run experiments without filing tickets for every toggle
ENGINEERS

Developers, SREs, DevOps — answer implementation questions instantly

Onboard to Unfamiliar Code in Minutes

"How does the recommendation engine work? Where's the caching layer? What triggers the nightly ETL?" Ramp up on systems you didn't build without scheduling sync meetings.

Ask: @context Explain how the fraud detection service makes decisions
Why it matters: Ship your first PR in hours, not weeks

Debug Production Issues Faster

"This Sentry error shows a nil pointer in OrderProcessor. What's calling this method? Where should the guard be?" Trace errors through the stack without grep gymnastics.

Ask: @context Why would OrderProcessor#finalize raise NoMethodError on user.email?
Why it matters: Reduce MTTR by finding root causes instantly

Answer Code Review Questions Async

"Why does this service use a custom serializer instead of ActiveModel? Is that intentional?" Let the codebase explain historical decisions so you don't have to.

Ask: @context Why do we use Sidekiq for webhooks instead of ActiveJob?
Why it matters: Unblock code reviews across time zones
MONITORING

Monitor every request in real-time

See exactly what your team is asking and how Claude Code responds

Easy Debug Panel

Track all incoming messages and Claude Code responses in a single, consolidated panel. Perfect for monitoring usage, debugging issues, and understanding how your team interacts with your codebase.

Or turn off logging if you want nothing stored.

✓
Real-time Monitoring

See every request as it happens with live updates

✓
Filter by Source

Narrow down to Slack, API, or chat widget requests

✓
Full Request History

Review payload and response details for every interaction

✓
Status Tracking

Monitor pending, processing, completed, and failed requests

Events Dashboard
All Processors
Timestamp
Source
Status
Request
Response
2025-01-19 14:32:15
Slack
Success
How do we handle webhook retries?
Webhook retries are handled in app/jobs/webhook...
2025-01-19 14:28:42
REST API
Success
Analyze Sentry error #4521
The error occurs in user_controller.rb:45 when...
2025-01-19 14:26:18
Chat Widget
Processing
Where is user authentication implemented?
—
CONTEXT

Make your chat responses smarter

Enrich the context for more accurate, personalized answers

1

Wrap Customer Information in Data Attributes

Pass rich contextual data before calling the embedded chat widget. Include customer details, pricing tiers, transaction history, feature flags, and system-level data.

data-customer-id="12345"
data-account-tier="enterprise"
data-pricing-plan="pro-annual"
data-user-role="admin"
2

Call Dynamic API Endpoints

Connect to your internal APIs to fetch real-time data. Pull in live metrics, user activity, or system status to answer questions with current information.

POST /api/v1/query
{
  "context_apis": [
    "https://api.yourapp.com/metrics"
  ]
}
3

Upload Static Documents

Supplement your codebase with additional documentation. Upload PDFs, text files, API specs, or runbooks to give Claude Code a complete picture.

PDFs Text Files API Specs Runbooks
PRICING

Simple, transparent pricing

No per-seat fees. No usage caps. Just pick your deployment.

Shared Server

Shared server hosted by cab.dev

$ 100 /month
  • ✓ Shared infrastructure
  • ✓ Standard support
  • ✓ Up to 10,000 API calls/month
  • ✓ Community access
  • ✓ SSL encryption
Get Started
Most Popular

Dedicated Server

Dedicated server instance hosted by cab.dev

$ 500 /month
  • ✓ Dedicated infrastructure
  • ✓ Priority support
  • ✓ Unlimited API calls
  • ✓ Custom domain
  • ✓ Advanced analytics
  • ✓ 99.9% SLA
Get Started

Self-Hosted

Docker image for self-hosting

$ 1000 /month
  • ✓ Docker image license
  • ✓ Self-hosted on your infrastructure
  • ✓ Full data ownership
  • ✓ Priority support
  • ✓ Custom integrations
  • ✓ White-label option
  • ✓ Unlimited usage
Get Started

Bring your own API key. You pay Anthropic/AWS/Google directly for LLM usage. We don't mark up token costs.

GET STARTED

Stop being the bottleneck.
Let your code answer questions.

2 minutes to set up. Your API key. Zero markup on LLM costs.

✓ Free to start ✓ No credit card ✓ Works in 2 minutes
Start Free → Technical due diligence without the engineering dependency.
$ context setup
Connecting to GitHub...
✓ Repository linked
Installing Slack app...
✓ Bot added to #engineering
Indexing codebase...
✓ Ready to answer questions
$ █