Now in early access — connect repos free

Every line of code
knows its origin.

Spectre tags every line across Copilot, Cursor, and Claude Code — AI-generated, human-written, or mixed. See which tools produce code that actually ships, and prove the value of your AI investment.

Get Early Access Book a Demo

47%

Avg AI code in new PRs

< 30s

Per PR analysis

19%

Orgs with full AI visibility

@@ -12,4 +12,18 @@

12 func handleRequest(ctx context.Context, req *http.Request) {

13- return processLegacy(req)AI

14+ if err := validateInput(req); err != nil {AI

15+ return nil, fmt.Errorf("validation failed: %w", err)AI

16+ }AI

17+ // TODO: add rate limiting herehuman

18+ result, err := processRequest(ctx, req)AI

19+ return result, nilAI

Legend:+ AI added- AI removed+ Human added- Human removed

The $3.5B blind spot

The AI code assistant market hit $3.5B in 2025. Engineering orgs spend $200K-$500K/year on these tools. Yet nobody can answer the most basic question.

$500K+/yr

Average enterprise AI tool spend with zero visibility into real ROI

19%

Of organizations have full visibility into where they use AI in code

1.7x

More issues in AI-assisted code vs human-written — without governance

Acceptance rate has zero correlation with what actually lands in production

Product Preview

See it in action

AI attribution, governance metrics, and tool ROI — all in one place.

AI Code Attribution

Across 12 repositories

30 days

0%AI Code

AI-Generated

0%+4.2%

24,180 lines

Mixed

0%+0.8%

3,210 lines

Human-Written

0%-5.0%

36,420 lines

AI Lines by Tool

#1GitHub Copilot

#2Cursor

#3Claude Code

#4Windsurf

#5Cline

Executive Dashboard

DORA + AI governance metrics

Q2 2026

Deploy Frequency

0.0

/day

+12%

AI Survival Rate

0.0

+3.2%

Change Failure Rate

0.0

-2.1%

AI Incident Rate

0.0

-0.8%

DXI Score

AI ROI

Shadow AI

AI Adoption Trend

+14% this quarter

JanFebMarAprMayJunJulAugSepOctNovDec

One pane of glass for all your AI tools

Copilot's analytics only see Copilot. Cursor's dashboard only sees Cursor. Spectre sees everything — across every tool, every repo, every team.

Cross-Tool Visibility

One dashboard for Copilot, Cursor, Claude Code, and Windsurf. See which tool produces the best code for your codebase — not just acceptance rates.

Code Survival Rate

Track which AI code ships to production vs gets rewritten in review. The metric that actually tells you if AI is helping or creating churn.

ROI Calculator

Map tool spend to landed code value. Factor in survival rate, defect rate, and hourly cost — in a format ready for board presentations.

Line-Level Attribution

Every line tagged — AI-generated, human-written, or mixed. Down to the agent, model, and prompt. Survives merge, rebase, and squash.

Policy Engine

Declarative rules: block PRs above 80% AI in auth/, require security review, fail status checks. Enforce governance automatically.

Compliance Reports

SOC 2, EU AI Act, and custom frameworks. One-click audit trails and AI code disclosure reports for regulated industries.

Integrations

Works with your stack

Hooks into every AI coding tool and connects to your entire DevOps stack.

Claude Code

Cursor

GitHub Copilot

Windsurf

Antigravity

GitHub

GitLab

Bitbucket

Slack

Microsoft Teams

Jira

Linear

PagerDuty

SonarQube

GitHub Actions

Datadog

Grafana

ServiceNow

Built for every stakeholder

VP of Engineering

Justify AI spend to the board

“The board asks what we get for $500K/year in AI tools. Spectre gives me the answer — with real landed-code data, not vendor-inflated accept rates.”

Security Engineer

Route AI code through stricter review

“AI code has different vulnerability patterns. Spectre lets me apply stricter review rules to exactly those lines — automatically, on every PR.”

Individual Developer

Understand your AI usage patterns

“I can see which prompts produce code that survives review vs gets rewritten. It has made me dramatically better at working with AI.”

Governance as code

Define policies in YAML. Run the CLI. Get instant visibility. Everything stays in your repo.

# Enforce review on high-AI auth code

rules:

- name: auth-ai-review

trigger:

ai_percentage: > 80%

paths:

- "src/auth/**"

- "src/crypto/**"

action: require_security_review

notify: slack:#security

# Initialize in any repo

$ spectre init

✓ Connected to GitHub

✓ Git hooks installed

✓ Policy loaded (2 rules)

✓ Dashboard linked

Ready. AI attribution will be tracked

on every commit automatically.

# Check a PR

$ spectre analyze --pr 312

⬤ AI: 67% ⬤ Human: 28% ⬤ Mixed: 5%

Trusted By

Our Customers