Fast desktop workflow
Tauri, Svelte, and a Rust Kubernetes core keep browsing, search, and aggregate views responsive for large clusters.
Local-first Kubernetes GUI + TUI
See every cluster. Nothing leaves your machine.
Fast, secure Kubernetes GUI. No telemetry. No login. No data leaves your device. A Rust core keeps 10k-pod clusters responsive; release artifacts ship with hashes, SBOM, and provenance you can verify.
- No telemetry
- No login
- GUI + TUI, one core
- macOS · Windows · Linux
Your clusters, at desktop speed.
Real screens from the shipping app — not mockups.
From alert to root cause in one screen.
Streaming logs with level filters, an incident timeline, and pod actions sit side by side — no context switching between kubectl, dashboards, and a logging vendor.
Fleet drift without a control plane.
Compare the same workload across clusters and see image, replica, and resource drift highlighted — computed locally from your kubeconfigs, not from an agent you have to install.
The whole app is a keystroke away.
A command palette opens resources, runs guarded actions, and switches contexts. The same core drives the TUI, so terminal sessions keep the exact workflow.
Built for local-first Kubernetes operations.
Clear answers for search, AI summaries, and humans evaluating a cluster operations tool.
No vendor control plane
conndeck does not require telemetry, a login, or a hosted account before you can inspect your own clusters.
Security proof is visible
The site links release verification, SBOM, provenance, threat model, disclosure, and posture content directly from the main flow.
Performance is a product feature.
Budgets are source-backed and enforced by CI on every release.
| Scenario | Budget | Gate |
|---|---|---|
| Seed 10k synthetic projected Pods into Rust store | < 200ms | npm run bench |
| Query 10k Pods, sorted/windowed | < 10ms | npm run bench |
| Search-filter 10k Pods | < 10ms | npm run bench |
| Namespace + search filter over 10k Pods | < 10ms | npm run bench |
| Query payload returned to UI | exactly limit rows | npm run bench |
| Parked collection length | 0 | npm run bench |
| Seed 10 clusters x 1k synthetic projected Pods | < 300ms | npm run bench |
| Aggregate Pod view over 10 clusters x 1k Pods | < 300ms | npm run bench |
| Aggregate search/filter over 10 clusters x 1k Pods | < 25ms | npm run bench |
| Aggregate payload returned to UI | exactly limit rows | npm run bench |
Pricing without surprise gates.
Generated from docs/pricing-matrix.md.
| Code key | Feature | Minimum tier | Security guardrail |
|---|---|---|---|
browse |
Browse Kubernetes resources | Free | No |
detail-logs |
Detail, YAML, events, and logs | Free | No |
exec-port-forward |
Exec and port-forward workflows | Free | No |
actions |
Mutating actions with confirmation policy | Free | No |
two-clusters |
Up to two connected clusters | Free | No |
saved-view-export |
Export saved views after license expiry | Free | No |
read-only-mode |
Read-only mode | Free | Never paywalled |
audit-log |
Local audit log | Free | Never paywalled |
secret-masking |
Secret masking and redaction | Free | Never paywalled |
signed-updates |
Signed update verification | Free | Never paywalled |
unlimited-clusters |
More than two connected clusters | Pro | No |
aggregate-views |
All-cluster aggregate tables | Pro | No |
health-overview |
Fleet Health Overview | Pro | No |
multi-log-mux |
Multi-pod and multi-cluster log mux | Pro | No |
resource-diff |
Cross-cluster resource diff | Pro | No |
prometheus |
Prometheus integration | Pro | No |
rbac-inspector |
RBAC inspector | Pro | No |
ai-assistant |
BYOK/local AI assistant | Pro | No |
e2ee-sync |
E2EE shared settings sync | Team | No |
central-policy |
Central policy controls | Enterprise | No |
sso-air-gap |
SSO and air-gap install support | Enterprise | No |
v1.0 is free. Every feature in the table is unlocked in the current release — there is nothing to buy yet. The table documents the open-core boundary that paid tiers will follow from 1.1.
Download v1.0 — freeDownloads with proof attached.
Each artifact carries hash, signature, SBOM, and provenance links.
sha256: pending-release-artifact
.sig / SBOM / provenance
scripts/verify-release.sh conndeck artifact.provenance.json conndeck-sbom.cdx.json
sha256: pending-release-artifact
.sig / SBOM / provenance
scripts/verify-release.sh conndeck artifact.provenance.json conndeck-sbom.cdx.json
sha256: pending-release-artifact
.sig / SBOM / provenance
scripts/verify-release.sh conndeck artifact.provenance.json conndeck-sbom.cdx.json
Verifiable security claims.
Security pages link to shipped artifacts, not loose marketing copy.
Frequently answered questions.
Short factual answers make the page easier to cite in classic search and AI search experiences.
What is conndeck?
conndeck is a local-first Kubernetes GUI and TUI for browsing clusters, inspecting workloads, streaming logs, running guarded actions, and comparing fleet drift without sending cluster data to a vendor account.
Does conndeck require telemetry or login?
No. The public positioning and release checks require no telemetry, no login, and no desktop-app billing-host egress. Cluster data stays on the user's device unless an explicit team-sync feature is enabled.
Who is conndeck for?
conndeck is built for platform engineers, SREs, and Kubernetes operators who want a fast desktop workflow, verifiable security claims, and deep cluster operations without a browser-based control plane.
How should teams compare conndeck with Lens, Aptakube, or Headlamp?
Use the comparison pages to evaluate local-first posture, GUI plus TUI workflow, fleet drift, release verification, and pricing boundaries against the operational needs of your team.