You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
[Coverage Report] Test Coverage Report — 2026-10-11
#9815
src/agent-entrypoint-output.test.ts
src/agent-helper-staging.test.ts
src/agent-runtime-start.test.ts
src/agent-runtime-start.ts
src/api-proxy-config-domains.test.ts
src/api-proxy-config-domains.ts
src/api-proxy-config-validation.test.ts
src/api-proxy-config-validation.ts
src/api-proxy-config-warnings.test.ts
src/api-proxy-config-warnings.ts
---
+export function resolveAgentRuntimeStartFile(workDir: string, proxyLogsDir?: string): string {
+export function readAgentRuntimeStartTimeMs(workDir: string, proxyLogsDir?: string): number | undefined {
+export function ensureAgentRuntimeStartMarker(workDir: string, proxyLogsDir?: string): number | undefined {
+export function assertCloudHypervisorEnclaveExportsIsolated(
+export function validateEnclavesStructure(
+export function resolveRunSandboxBackend(primaryRuntime?: string): RunSandboxBackend {
+export function validateRunSandboxBackend(
+export async function captureEnclaveStartupDiagnostics(
+export async function planNvxLiveMounts(
+export function appendMountArguments(
🔎 Notable Findings
All five security-critical files (host-iptables.ts, squid-config.ts, docker-manager.ts, domain-patterns.ts, cli.ts) are at or near full statement coverage; no regression is visible in the pre-computed data.
src/cli.ts has only 50% branch coverage, the weakest point on the security-critical list.
src/bounded-execution/finite-disclosure.ts has 11.42% branch coverage and 37.5% function coverage, the lowest branch figure in the report.
src/nvx/cleanup-registry.ts (42.8% statements) and src/bounded-execution/finite-cardinality.ts (46.03% statements) are the largest gaps overall. Neither is on the security-critical list.
🎯 Recommendations
High: Add tests for src/cli.ts branch paths (currently 50%), focusing on error and flag-combination handling in the orchestration flow.
Medium: Add tests for src/nvx/cleanup-registry.ts and src/bounded-execution/finite-cardinality.ts, targeting the uncovered branches listed in the coverage data.
Low: Raise branch coverage in src/bounded-execution/finite-disclosure.ts (11.42%) and src/microvm/network-reservation.ts (54.86%).
Generated by test-coverage-reporter workflow. Trigger: push
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
📊 Test Coverage Report — 2026-10-11
Overall Coverage
🛡️ Security-Critical Path Status
📋 Coverage Table
🔧 Function Audit
📅 Recent Source Changes (last 7 days)
🔎 Notable Findings
host-iptables.ts,squid-config.ts,docker-manager.ts,domain-patterns.ts,cli.ts) are at or near full statement coverage; no regression is visible in the pre-computed data.src/cli.tshas only 50% branch coverage, the weakest point on the security-critical list.src/bounded-execution/finite-disclosure.tshas 11.42% branch coverage and 37.5% function coverage, the lowest branch figure in the report.src/nvx/cleanup-registry.ts(42.8% statements) andsrc/bounded-execution/finite-cardinality.ts(46.03% statements) are the largest gaps overall. Neither is on the security-critical list.🎯 Recommendations
src/cli.tsbranch paths (currently 50%), focusing on error and flag-combination handling in the orchestration flow.src/nvx/cleanup-registry.tsandsrc/bounded-execution/finite-cardinality.ts, targeting the uncovered branches listed in the coverage data.src/bounded-execution/finite-disclosure.ts(11.42%) andsrc/microvm/network-reservation.ts(54.86%).Generated by test-coverage-reporter workflow. Trigger:
pushAll reactions