nightshift: test-gap — 5 source files (557 lines) have zero test coverage

[nightshift-micr]

nightshift: test-gap — Test Coverage Gap Analysis

Repo: Microck/tailstick
Language: Go 1.25.6
Analyzed: 2026-04-30

---

Summary

Tailstick has 6 test files covering 5 packages with ~30 test functions. However, 5 source files (2,432 lines, 38% of the codebase) have zero test coverage. Critical gaps exist in state persistence, platform detection, CLI command dispatch, and logging.

---

Coverage Map

Package	Source File	Lines	Test File	Status
app	workflow.go	755	workflow_test.go ✅	15 tests — good
tailscale	client.go	270	client_test.go ✅	7 tests — good
gui	server.go	247	server_test.go ✅	4 tests — good
config	config.go	116	config_test.go ✅	1 test — minimal
crypto	secret.go	128	secret_test.go ✅	2 tests — adequate
state	store.go	82	❌ None	0 tests
platform	platform.go	175	❌ None	0 tests
platform	exec.go	41	❌ None	0 tests
app	cli.go	196	❌ None	0 tests
logging	logger.go	63	❌ None	0 tests
model	types.go	133	N/A (types only)	Low priority

Coverage Statistics

• Files with tests: 5 / 11 source files (45%)
• Lines with tests: ~1,516 / ~2,432 (62% by line count)
• Untested lines: 557 across 5 files

---

Priority Gaps

🔴 HIGH — internal/state/store.go (82 lines, 0 tests)

Functions untested:

• Load(path string) — reads JSON state file, handles missing/corrupt files
• Save(path string, state LocalState) — writes state file, creates parent dirs
• UpsertRecord(state LocalState, rec LeaseRecord) — merges lease records
• AppendAudit(path string, entry AuditEntry) — appends NDJSON audit entries

Why critical: State persistence is the core of lease management. A bug in UpsertRecord could silently lose lease records, causing orphaned Tailscale devices. AppendAudit failures could lose audit trails required for compliance.

Recommended tests:

• TestLoadReturnsDefaultOnMissingFile
• TestLoadReturnsErrorOnCorruptJSON
• TestSaveCreatesParentDirectories
• TestUpsertRecordAddsNewRecord
• TestUpsertRecordUpdatesExistingByID
• TestAppendAuditCreatesFileIfMissing
• TestAppendAuditAppendsToExistingFile

---

🔴 HIGH — internal/platform/platform.go (175 lines, 0 tests)

Functions untested:

• Detect() — detects OS, hostname, boot ID for lease tracking
• IsLinux(), IsWindows() — platform checks
• StatePath(), LogPath(), LocalSecretPath(), AgentBinaryPath() — path resolution
• RequireSupportedLinux() — validates distro support
• IsElevated() — privilege check
• ElevationHint(exePath, args) — generates sudo/runas messages
• sanitizeHost(in string) — hostname sanitization

Why critical: Platform detection errors could cause leases to run with wrong paths or fail privilege checks. sanitizeHost is a security-sensitive input sanitization function.

Recommended tests:

• TestSanitizeHostStripsSpecialChars
• TestDetectSetsCorrectOS
• TestStatePathPlatformSpecific
• TestIsElevatedCurrentContext
• TestRequireSupportedLinuxRejectsUnknownDistro

---

🟡 MEDIUM — internal/app/cli.go (196 lines, 0 tests)

Functions untested:

• RunCLI(args, Runtime) — main CLI dispatch (routes to run/agent/cleanup/version)
• runEnroll(args, Runtime) — enrollment flow parsing
• runAgent(args, Runtime) — agent mode parsing
• runCleanup(args, Runtime) — cleanup flow parsing
• interruptContext() — signal handling
• printHelp() — help output

Why important: CLI dispatch is the user-facing entry point. Flag parsing bugs here surface as confusing errors. The interruptContext function handles graceful shutdown.

Recommended tests:

• TestRunCLIDispatchesRunCommand
• TestRunCLIDispatchesAgentCommand
• TestRunCLIDispatchesCleanupCommand
• TestRunCLIUnknownCommandReturnsError
• TestRunCLIPrintsHelp

---

🟡 MEDIUM — internal/platform/exec.go (41 lines, 0 tests)

Functions untested:

• Runner.Run(ctx, args) — executes commands with context cancellation
• Runner.RunWithTimeout(timeout, args) — timeout-wrapped execution

Why important: Command execution is security-sensitive. These wrap os/exec calls that run Tailscale commands on the system. Context cancellation and timeout behavior should be verified.

Recommended tests:

• TestRunnerRunExecutesCommand
• TestRunnerRunRespectsContextCancellation
• TestRunnerRunWithTimeoutExpires

---

🟢 LOW — internal/logging/logger.go (63 lines, 0 tests)

Functions untested:

• New(path) — creates log file
• Info/Error — formatted logging
• Close() — file cleanup

Why low: Logging is non-critical. Failures here don't affect lease correctness.

---

Existing Test Quality Notes

• config_test.go has only 1 test — should cover more config loading scenarios (missing fields, invalid JSON, env var precedence)
• crypto/secret_test.go covers encrypt/decrypt but not edge cases (empty password, very long passwords, key file permission errors)

---

Recommended Priority Order

1. state/store.go — highest impact, data integrity
2. platform/platform.go — security-sensitive input handling
3. app/cli.go — user-facing correctness
4. platform/exec.go — execution safety
5. logging/logger.go — nice to have

---

Generated by nightshift — test-gap analysis