Skip to content

Latest commit

 

History

History
53 lines (43 loc) · 6.65 KB

File metadata and controls

53 lines (43 loc) · 6.65 KB

Coverage Omit Ledger

This document maintains the governance ledger for all files and directories omitted from test coverage reporting in pyproject.toml under [tool.coverage.run].omit.

The ledger is not a permission slip for untested code. An omit pattern is acceptable only when it has an owner surface, a reason, a risk rating, a return milestone, and at least one smoke-test candidate or current adjacent evidence. scripts/validate_docs_consistency.py checks that the omit patterns listed here match pyproject.toml.

Ledger

Omit Pattern Owner Reason Risk Expected Return Milestone Smoke-Test Candidate
teaagent/tournament/* Swarm Consensus Team Optional tournament mechanics for federated evaluation paths. Medium: coordination scoring or ranking drift can silently reduce swarm quality. Phase 2: add deterministic tournament scenario coverage before production swarm positioning. tests/test_tournament.py, tests/test_tournament_parallel_executor.py, tests/acceptance/test_consensus_flow.py
teaagent/validation/* Governance Team Static validation rule definitions and schema-like helpers are easier to assert through rule outcomes than raw line coverage. Low: static rule drift is usually caught by validators, but missing rule coverage can weaken governance claims. Phase 1: add rule parsing and invalid-rule fixture coverage. tests/test_validation.py, tests/test_tranche_b_governance.py
teaagent/workflow_engine.py Subagents Team Background workflow graph executor with async/state-machine behavior. Medium: lock, state transition, or retry bugs can break long-running agent workflows. Phase 1: add headless workflow transition tests with persistence fixtures. tests/test_phase5_workflow_engine.py, tests/test_phase4_workflow_engine.py
teaagent/webhook_sink.py Platform Team External webhook posting is integration-shaped and better checked with a local HTTP target. Low: delivery regressions affect observability more than core safety. Phase 2: add telemetry mock validation and retry/backoff checks. tests/acceptance/test_webhook_audit_flow.py, tests/integration/test_webhook_sink.py
teaagent/wasm_runtime.py Sandbox Team Optional WASM engine depends on optional wasmer availability and platform support. Medium: sandbox execution or memory isolation bugs can affect optional untrusted-code paths. Phase 1: add conditional WASM engine integration tests that skip only when runtime is unavailable. tests/test_wasm_runtime.py
teaagent/wasm_skill.py Sandbox Team WASM-packaged skill manifest and invocation contract logic is tied to optional skill build flows. Medium: package validation drift can allow invalid or unsafe skill artifacts. Phase 1: add manifest validation and compilation-mock coverage. tests/test_phase6_remaining_features.py, tests/test_skill_executor.py
teaagent/tsb_format.py Governance Team Tool signing bundle serialization spans filesystem, audit evidence, and manifest verification. Medium: bundle parsing differences can weaken supply-chain evidence. Phase 1: keep unit coverage plus lifecycle integration before release packaging. tests/test_tsb_format.py, tests/integration/test_tsb_lifecycle.py
teaagent/workspace_tools/builder.py Workspace Team Builder wraps local workspace tool registration and setup paths. Low: local setup wrapper errors are likely visible quickly, but can confuse first-run UX. Phase 1: add local tool setup smoke tests to coverage suite. tests/test_workspace_tools.py::WorkspaceToolConfigTests::test_builder_with_workspace_tools
teaagent/workspace_tools/_git.py Workspace Team Host git integration is safer to test with temp repos and mocks than global coverage pressure. Medium: argument mistakes can affect commit, stash, and recovery workflows. Phase 1: expand mocked git harness and temp-repo lifecycle tests. tests/test_git_tools.py, tests/test_lore_commit_formatter.py
teaagent/workspace_tools/_config.py Workspace Team Workspace defaults and ignore matching are configuration helpers used through higher-level tools. Low: config parsing errors are bounded to workspace-tool setup. Phase 1: add TOML parser coverage and invalid config fixtures. tests/test_workspace_defaults_toml.py, tests/test_tool_dependency_injection.py
teaagent/browser_tools.py Workspace Team Optional browser automation tools rely on Playwright-like drivers and read-only tool registration. Medium: browser driver mismatch or mutating action exposure can violate operator expectations. Phase 1: add Playwright headless tests when optional extra is available. tests/test_browser_tools.py, tests/acceptance/test_browser_tools_integration_flow.py
teaagent/cli/_handlers/_cost.py Core CLI Team Legacy standalone cost command is superseded by chat/TUI controller-backed cost reporting. Low: main risk is stale UX or duplicate cost truth. Phase 0: deprecate, delete, or add direct handler smoke before release. tests/acceptance/test_cost_tracking_flow.py, tests/test_cli_ergonomics_handlers.py

Omissions Removed (Phase 3 Round)

The following patterns were removed from the omit list in pyproject.toml and this ledger because adequate test coverage now exists:

Previous Omit Pattern Justification
teaagent/tui/* 20 TUI command-path tests added in tests/tui/ covering command truth, typo confirmation, and undo scope. Partial coverage is sufficient to remove the omit.
teaagent/vote_relay.py Tests exist in tests/test_strategic_features.py, tests/test_surface_auth_hardening.py, tests/acceptance/test_security_vote_relay_flow.py, tests/test_remediation_p1_p2.py, and tests/test_analysis_followups.py.
teaagent/tls_server.py Acceptance tests in tests/acceptance/test_security_tls_server_flow.py cover both functions (build_server_ssl_context, wrap_server_socket) with self-signed certs.
teaagent/cli/_handlers/_control_plane.py Dedicated test file tests/test_cli_handlers_control_plane.py (5 tests, 160 lines) covers serve command with mocks.

Re-Entry Rules

  • New omit patterns require a same-commit ledger row and a linked smoke-test candidate.
  • Security-sensitive omit patterns (tls_server, wasm_runtime, wasm_skill, browser tools, approval/control-plane surfaces) require a central risk or roadmap link when their risk is High or Critical.
  • Removing an omit pattern is allowed only when the relevant tests run under the normal coverage command, not only in a hand-run optional environment.
  • Legacy surfaces should be deleted or deprecated before adding more indirect tests around them.