Features
Everything you need to ship AI-assisted code with confidence.
Structured plans, complexity alerts, real-time dashboards, and deep integrations — the trust layer between your team and the AI writing your code.
Agent-assisted plans with human approval
Define intent before AI writes a single line. Structured plans translate your goals into safe, bounded execution that agents follow — and humans approve.
- Machine-readable specs keep agents on track
- Review and refine plans before code is touched
- Explicit scope prevents agent drift and surprises
Plan snapshot
Goal
Add audit logging to user settings service without downtime.
Scope
- Update `services/settings` handlers
- Add redact helper in `lib/logging`
- No DB schema changes
Complexity metrics that surface risk early
Automated alerts fire when AI-generated code introduces complexity regressions, high coupling, or other code smells — before they become tech debt.
- Nesting depth, branch count, and LOC monitoring
- Configurable severity: Info, Warning, Critical
- Actionable guidance, not noisy false positives
Alert snapshot
Trigger
Branch count exceeded medium threshold.
Severity
Metrics, health scores, and trends at a glance
A centralized view of your codebase health. Track complexity trends, monitor alert patterns, and measure the impact of AI-assisted changes over time.
- Repository health scores across your org
- Historical trend tracking for key metrics
- Plan completion and alert resolution analytics
Dashboard
Health Score
87/100
Active Alerts
3
Plans Completed
24
Files Scanned
1.2k
Auto-scan on merge, zero config needed
Connect your repos and Spec This scans every merge automatically. Discover new repositories, track changes across branches, and surface issues where your team already works.
- One-click GitHub App installation
- Automatic scanning on push and merge
- Repository discovery and org-wide rollout
GitHub Integration
acme/web-app
Connected · Auto-scan on merge
acme/api-server
Connected · 12 scans this week
Local scanning and AI-native integrations
Run scans locally via the CLI before you push. Connect Claude, Cursor, or Codex through the MCP server so AI agents work within your plan boundaries automatically.
- Scan locally with a single command
- MCP server for Claude, Cursor, and Codex
- Results sync to your dashboard in real time
CLI & MCP
$ npx spec-this scan ./src
✓ 142 files scanned
✓ 3 alerts generated
✓ Results synced to dashboard
$ npx spec-this mcp start
MCP server ready on localhost:3100
Ready to Ship with Confidence?
Join teams that trust their AI-assisted code. Start free, no credit card required.