Teams using Drape recover 4+ hours per developer per week and cut false-red CI runs by 90%.
Works with your stack
Built for the agentic coding era — because we built the brain
AI coding agents generate code at scale — but when tests fail, they're flying blind. Drape is the CI Brain: it auto-triages every red build (real bug, flake, or infra), suppresses known flakes, and exposes structured failure context via API and MCP — so agents fix what matters without punting back to a human. Run burn-in and bisect from chat, PR comment, or API.
This is what your dashboard actually looks like
Real components from the Drape product — not mockups.
CI Time Wasted This Week
Recent Changes
Test Health at a Glance
| Test | Status | Last 10 Runs |
|---|---|---|
test_user_login |
Flaky |
|
test_checkout_flow |
Stable |
|
test_api_timeout |
Flaky |
|
test_search_results |
Recovering |
|
Top Failure Clusters
| Pattern | Occurrences | Jobs |
|---|---|---|
| Connection timeout in test_api_* | 23 | 5 |
| OOM killed in build step | 12 | 3 |
| Fixture teardown race condition NEW | 8 | 2 |
From broken build to shipped fix — automatically
Drape doesn't just show you what's wrong. It suppresses false negatives, blocks coverage regressions, and gives your team and AI agents the context to decide what to fix next.
Discover
Know in seconds whether a failure is a real bug, a flake, or a CI environment issue
Flaky Test Detection
Automatically score and flag flaky tests so your team stops chasing phantom failures.
| test_login | ||||||
|---|---|---|---|---|---|---|
| test_checkout | ||||||
| test_api | ||||||
| test_export |
Test Health Grid
Every test, every commit. Scattered red means flaky. Solid columns mean a CI environment issue. Patterns jump out instantly.
AI Failure Classification
Automatically triage every failure — real bug, flaky test, or CI environment issue — so developers stop wasting time reading logs.
Fix
Cluster failures by root cause and ship one PR instead of fifteen rerun loops
Failure Clusters
15 failing tests, 3 root causes. Drape clusters by stack trace and points you at the file and line.
PR Comment Bot
Drape posts triage context directly on the PR — no opening a separate tool.
Smart Suppression
Quarantine known flakes with a deadline. Drape escalates anything still suppressed past your SLA.
Prevent
Automated guardrails that block regressions at the PR
Coverage That Only Goes Forward
Enforce line-level coverage. Block PRs that drop coverage on changed files — no more gaming percentage targets.
Drape · PR #142
⚠️ 3 lines lost coverage in middleware.py
🆕 New flake: test_login_timeout
🐌 test_export 3.2x slower
⏱️ Merge cycle: 4.2h Elite
PR Insights
Drape posts actionable context on every PR: new flakes, coverage regressions, duration spikes, and DORA merge cycle metrics — before the code merges.
Container CVE Tracking
Ingest SARIF scans from Grype or Trivy. Block PRs that introduce new critical vulnerabilities. Enforce SLA-based fix timelines by severity.
# Install via Homebrew, Docker, or Nix
brew install drape-io/tap/drape
drape upload tests results.xml --wait
CLI, GitHub Action, or raw HTTP.
Standard JUnit XML, SARIF, Cobertura, LCOV, and Go coverage. Use the Drape CLI, the GitHub Action, or any HTTP client. No SDK to embed in your code, no vendor lock-in.
Reliable CI in 5 minutes
Connect your repository
Install the GitHub App or add a GitLab webhook. One click.
Upload your CI data
Use the GitHub Action or the Drape CLI to upload test results, coverage, security scans, and lint reports. Standard formats — JUnit XML, SARIF, Cobertura, LCOV — no vendor lock-in.
# .github/workflows/ci.yml
- name: Upload test results
uses: drape-io/drape-action@v2
with:
command: tests
file: results.xml
api-key: ${{ secrets.DRAPE_API_KEY }}
Drape takes action
Flaky tests get suppressed. Coverage regressions and new CVEs get blocked. Lint violations get tracked. The GitHub Action posts rich PR comments with results. AI agents get the context to help fix what matters. Your team ships.