🔑 Coming Soon

1Password Integration

Manage and rotate secrets directly from your 1Password vaults — no copy-paste, no plain-text keys in Konstruct's catalog.

  • Browse vaults and items via the 1Password CLI bridge
  • Reference secrets by path — values are resolved at runtime, never stored
  • Works alongside local and environment-variable secrets
  • Requires Konstruct Pro
🔍 Coming Soon

Dependency Visibility

See every dependency across all of your repositories in one place — even multi-repo projects with sub-projects in different languages.

  • Unified dependency view across all repos and sub-projects
  • Detects outdated packages with available upgrade paths
  • Works across multiple languages in the same project
  • Integrated directly into your project's code graph
🛡️ Coming Soon

Vulnerability Analysis

Find out which of your dependencies have open CVEs — before they become a problem.

  • Cross-references your full dependency tree against known CVE databases
  • Highlights severity and affected versions at a glance
  • Works across all repositories and languages in a project
  • Pairs with Dependency Remediation for one-click fixes
📦 Coming Soon

Dependency Remediation

Go beyond visibility — automatically upgrade outdated or vulnerable dependencies with agent-assisted fix diffs, in-depth impact analysis, and guided remediation across all your repositories.

  • Upgrade diffs generated and reviewed before any changes are applied
  • In-depth impact analysis across your codebase for each upgrade
  • Agent-assisted fixes for breaking changes introduced by upgrades
  • Works across multi-repo projects in multiple languages
🏗️ Coming Soon

Build Sandbox

Keep agent verification inside build steps your team defines and checks into git. Instead of improvising shell commands, agents run only the lint, test, and build work you have approved.

  • Define build targets in a per-repo manifest checked into git
  • Buildpacks automatically install and isolate the right toolchain per project
  • Chain prerequisites so lint runs before test, test before build—automatically
  • Block arbitrary shell commands during agent verification runs
🔀 Coming Soon

GitHub PR Management

View and manage pull requests across all your project repositories directly inside Konstruct. Get AI-driven review guidance highlighting what to look for and overall quality signals.

  • PR list and detail view across multi-repo projects
  • Konstruct’s analysis of code quality and change impact
  • Checklist-style review guidance surfaced by AI
🔒 Coming Soon

Security Reviews

Iterative, agent-driven security reviews of your codebase with ranked findings and prioritized mitigation checklists you can work through with AI assistance.

  • Multi-round review sessions with persistent findings
  • Severity-ranked vulnerability list with context
  • Agent-guided remediation steps and code suggestions
Coming Soon

RunPod GPU Integration

Manage RunPod GPU instances directly from Konstruct—start or stop them on demand to control spend and run large local models without leaving your workflow.

  • Start and stop GPU pods on demand
  • Cost visibility and spend controls
  • Seamless routing to RunPod-hosted LLMs as an AI provider
💰 Coming Soon

Token Savings

Reduce LLM token usage without sacrificing context quality through deduplication, extraction, and minification of context passed to AI providers.

  • Deduplication — eliminate repeated content across context windows
  • Extraction — pull only the most relevant symbols and sections
  • Minification — compact context without losing meaningful signal
🖥️ Coming Soon

VM Management

Provision and manage virtual machines as Konstruct workspaces so your workspace daemon and agents run in an isolated environment separate from your laptop.

  • Create and connect to VMs as project workspaces
  • Agents run inside the VM—isolated from the host
  • Same workspace daemon experience as local development
🤖 Coming Soon

Multi-Agent Code Reviews

PR-style code reviews driven by multiple specialized agents—each focused on a different dimension like correctness, security, performance, and style—with consolidated threaded feedback.

  • Multiple review agents with distinct focuses
  • Threaded inline feedback on diffs
  • Consolidated review summary with action items
📊 Coming Soon

Code Quality Analyzer

Continuous code quality analysis combining algorithmic metrics and AI review to surface complexity hotspots, structural smells, and maintainability issues — especially useful for keeping AI-generated code in check.

  • Cyclomatic complexity measurement per function across TypeScript and Go
  • Function length, nesting depth, and structural smell detection
  • AI-driven review layered on top for deeper semantic insight
  • CI-gate support — fail builds when complexity thresholds are breached