dicklesworthstone__coding_agent_session_search/docs/test-coverage-audit.md

Test Coverage Audit: coding_agent_session_search (cass)

Date: 2026-01-12 Bead: coding_agent_session_search-vh1n

---

Executive Summary

• Total Tests: ~1,938 tests
  • Inline unit tests (src/): ~1,091
  • External tests (tests/): ~847
• Mock/Fake Usage: All mocks are test helpers within #[cfg(test)] modules - no external mock libraries
• Critical Gaps Identified: 3 high-risk modules with 0 or minimal inline tests

---

1. Test Classification by Level

Unit Tests (Single component/function isolation)

Location	Count	Notes
src/connectors/*.rs	447	Comprehensive - each connector has 14-57 tests
src/search/*.rs	277	Excellent coverage of query.rs (147 tests)
src/sources/*.rs	159	Good coverage of config, sync, provenance
src/ui/*.rs	108	tui.rs (89), data.rs (8), time_parser.rs (4)
src/pages/*.rs	44	fts.rs (15), size.rs (10), encrypt.rs (8)
src/storage/sqlite.rs	36	Storage layer coverage
src/indexer/mod.rs	20	Indexer logic
tests/connector_*.rs	266	Additional external connector tests
tests/storage.rs	44	Storage integration tests
tests/ui_*.rs	73	UI component tests
tests/search_*.rs	8	Search algorithm tests
Unit Subtotal	~1,482

Integration Tests (Multiple modules together)

File	Count	Coverage
tests/setup_workflow.rs	26	Setup wizard flow
tests/semantic_integration.rs	21	Semantic search pipeline
tests/pages_fts.rs	15	Pages full-text search
tests/multi_source_integration.rs	14	Multi-source handling
tests/pages_bundle.rs	12	Export bundling
tests/pages_export.rs	10	Export formats
tests/ssh_sync_integration.rs	10	SSH sync
tests/concurrent_search.rs	6	Concurrent access
tests/logging.rs	3	Log capture
tests/memory_tests.rs	3	Memory safety
tests/ssh_test_helper.rs	2	SSH utilities
Integration Subtotal	122

End-to-End Tests (Full application flow)

File	Count	Coverage
tests/cli_robot.rs	138	Robot mode CLI commands
tests/e2e_sources.rs	37	Source management
tests/e2e_filters.rs	22	Filter functionality
tests/e2e_cli_flows.rs	20	CLI workflows
tests/e2e_search_index.rs	15	Search + index
tests/e2e_multi_connector.rs	8	Multiple connectors
tests/install_scripts.rs	7	Installation
tests/cli_index.rs	6	Index commands
tests/perf_e2e.rs	5	Performance e2e
tests/watch_e2e.rs	4	Watch mode
tests/perf_proptest.rs	3	Property-based perf
tests/e2e_index_tui.rs	1	TUI indexing
tests/e2e_install_easy.rs	1	Easy install
E2E Subtotal	267

---

2. Coverage Gap Matrix by Module

Module	Inline Tests	External Tests	Gap Level	Priority
lib.rs	0	CLI covered by e2e	HIGH	P0
model/	0	None	HIGH	P1
encryption.rs	0	crypto_vectors (3)	MEDIUM	P2
bookmarks.rs	8	None in tests/	LOW	P3
export.rs	7	pages_export (10)	LOW	P3
update_check.rs	8	None	MEDIUM	P2
connectors/	447	266	GOOD	-
search/	277	29	GOOD	-
storage/	36	44	GOOD	-
ui/	108	73	GOOD	-
pages/	44	37	GOOD	-
sources/	159	122	GOOD	-
indexer/	20	5	LOW	P3

Critical Gap Details

lib.rs (0 tests) - P0 Critical

• Size: ~11,500 lines
• Risk: Main CLI entry point, arg parsing, command dispatch
• Contains: Error types, normalize_args(), execute_cli(), all command handlers
• Recommendation: Add unit tests for:
  • Argument normalization/auto-correction
  • Error type conversions
  • Individual command parsing

model/ (0 tests) - P1 High

• Files: mod.rs, types.rs
• Risk: Core data types (Agent, Conversation, Message, Snippet)
• Recommendation: Add tests for:
  • Serialization/deserialization
  • Type conversions
  • Default values

encryption.rs (0 inline tests) - P2 Medium

• External coverage: tests/crypto_vectors.rs (3 tests)
• Risk: Security-critical code
• Recommendation: Increase test coverage with:
  • More edge cases
  • Error path testing
  • Key derivation tests

---

3. Mock/Fake/Stub Analysis

Summary

• External mock libraries: NONE (no mockall, mockito, etc.)
• Internal test helpers: All within #[cfg(test)] modules

Identified Mock/Fake Usage

Acceptable Test Fixtures (No Replacement Needed)

File	Pattern	Purpose
tests/connector_claude.rs	mock-claude/ directory	Fixture directory for Claude connector tests
tests/parse_errors.rs	mock-claude/ directory	Same fixture pattern
tests/e2e_install_easy.rs	fake_bin/, fake binaries	Simulates installed tools for install tests
tests/semantic_integration.rs	fake_model.onnx	Simulates ML model file
tests/install_scripts.rs	Fake binary	Upgrade path testing

Internal Test Helpers (Within #[cfg(test)] - Acceptable)

File	Helper	Purpose
src/search/embedder.rs:190-224	MockEmbedder	Test embedder trait impl
src/ui/tui.rs:8976-9008	MockHit, MockPane	TUI selection testing
src/sources/install.rs:935-1106	mock_system_info(), mock_resources()	Install strategy tests
src/sources/index.rs:608-674	mock_probe_* functions	Host probe testing

Verdict

No prohibited mock usage found. All mock/fake patterns are:

1. Test fixture directories (not behavioral mocks)
2. Internal test helpers within #[cfg(test)] modules
3. Simple data fixtures for testing

---

4. High-Risk Untested Paths

Error Handling Paths

Location	Risk	Current Coverage
lib.rs error types	HIGH	E2E only
Storage migrations	MEDIUM	tests/storage.rs
Index corruption recovery	MEDIUM	Some in indexer tests
SSH connection failures	MEDIUM	ssh_sync_integration

Performance/Resource Paths

Path	Risk	Current Coverage
Large archive (>10K convos)	HIGH	perf_e2e (limited)
Memory pressure	MEDIUM	memory_tests (3)
Concurrent access	MEDIUM	concurrent_search (6)

Security Paths

Path	Risk	Current Coverage
AES-GCM encryption	HIGH	crypto_vectors (3)
Secret scanning	MEDIUM	secret_scan (3)
Nonce generation	MEDIUM	security_nonce (7)

---

5. Remediation Priority Ordering

P0 - Critical (Should block releases)

1. Add lib.rs unit tests - Argument parsing, error types, command dispatch
2. Add model/ unit tests - Core type serialization/conversion

P1 - High (Next sprint)

3. Expand encryption.rs tests - More vectors, error paths
4. Add update_check.rs integration tests - Version comparison, network errors
5. Large archive performance tests - 10K+ conversation benchmarks

P2 - Medium (Backlog)

6. Migration rollback tests - Storage schema downgrade paths
7. TUI interaction coverage - More keyboard/mouse event tests
8. Error message quality tests - User-facing error strings

P3 - Low (Nice to have)

9. Bookmarks module external tests
10. Indexer edge case coverage
11. Export format fidelity tests

---

6. File Links for Key Test Locations

Inline Tests (src/)

• src/search/query.rs - 147 tests (search engine)
• src/ui/tui.rs - 89 tests (TUI logic)
• src/connectors/amp.rs - 57 tests (Amp connector)
• src/connectors/mod.rs - 49 tests (connector framework)
• src/connectors/gemini.rs - 46 tests (Gemini connector)

External Tests (tests/)

• tests/cli_robot.rs - 138 tests (robot mode)
• tests/ui_snap.rs - 50 tests (UI snapshots)
• tests/storage.rs - 44 tests (storage layer)
• tests/connector_aider.rs - 39 tests (Aider connector)
• tests/e2e_sources.rs - 37 tests (source management)

Benchmarks (benches/)

• benches/bench_utils.rs - Performance benchmarks

---

7. Next Steps

1. Create beads for P0 remediation tasks
2. Add lib.rs test module with arg parsing tests
3. Add model/ serialization tests
4. Expand crypto test vectors
5. Schedule performance regression baseline

---

8. Connector Test Audit (bead mo6o)

Fixture Strategy Analysis

Real Fixture Files (tests/fixtures/):

Connector	Fixture Path	Format
aider	fixtures/aider/.aider.chat.history.md	Markdown
amp	fixtures/amp/thread-001.json	JSON
cline	fixtures/cline/task1/	JSON files
claude_code	fixtures/claude_code_real/projects/	JSONL
codex	fixtures/codex_real/sessions/	JSONL
gemini	fixtures/gemini/hash123/chats/	JSON
opencode	fixtures/opencode_json/	JSON
pi_agent	fixtures/pi_agent/sessions/	JSONL
multi_source	fixtures/multi_source/	JSONL

Inline Fixture Tests (no dedicated fixture files):

Connector	Inline Tests	Notes
chatgpt	30	Uses inline JSON in test code
cursor	43	Uses inline data in test code
factory	14	Uses inline data in test code

Mock/Fake Pattern Analysis

Acceptable Patterns Found:

• mock-claude/ - Directory name for test fixtures (NOT behavioral mock)
• TempDir::new() - Creates real temporary directories
• Inline JSONL/JSON literals - Mirror actual agent output formats

No Prohibited Patterns Found:

• No mockall/mockito library usage
• No trait mocks for Connector interface
• All tests use actual filesystem operations

Edge Case Coverage (parse_errors.rs, fs_errors.rs)

Category	Tests	Status
Invalid JSON syntax	4	Covered
Missing required fields	3	Covered
Wrong field types	2	Covered
Truncated files	2	Covered
Binary/null bytes	2	Covered
Non-existent paths	3	Covered
Permission denied	2	Covered (Unix)
Symlink handling	4	Covered (Unix)
Unicode paths	2	Covered
Empty directories	2	Covered

Verdict

Connector tests meet the requirements:

1. All use real fixture data that mirrors actual on-disk formats
2. No behavioral mocks - only test fixture directories
3. Edge cases comprehensively covered in parse_errors.rs and fs_errors.rs
4. Tests use actual filesystem paths via TempDir

Minor Improvements (Optional):

• Add dedicated fixture files for chatgpt, cursor, factory connectors
• Add cross-platform path handling tests for Windows

---

Generated by test coverage audit - beads vh1n, mo6o