Production unit-test evidence inventory — 2026-10-14

Preparation for Issue #496. Assessed source revision: 34af7f8226647c722899a5837eff7e887cf7fbe6 (origin/main at discovery). Inventory captured 2026-10-14 UTC. This is a historical snapshot; later changes require a fresh inventory.

Result and limits

All 62 production skills are inventoried below: 59 have non-test Python helper candidates, and three have none. The index declares unit_tests: not_verified for the 59 and not_applicable for the three. This inventory changes no verification declarations.

This is a file and policy inventory, not a completed semantic test audit. It does not satisfy any acceptance criterion of #496, prove missing coverage, or establish passed. The confirmed evidence gap is the conservative shared baseline rationale: no retained, helper-by-helper happy/boundary/error-path, fail-closed, and schema-conformance audit. Whether existing tests already meet those requirements is NOT_ASSESSED here. Tests were not executed to produce this table. File counts cannot establish test quality. Security review and instruction-contract review are also NOT_ASSESSED.

Discovery and triage rules

• Read the index, policy, file modes, and file contents from the exact source revision using git show / git ls-tree; untracked files, local caches, and the working tree cannot alter the snapshot.
• Include every status: production ID exactly once, sorted by ID. Scope is each skill’s scripts/ tree. Files outside that tree are not assessed as helper candidates.
• Count non-test .py files as helper candidates, including generated clients and library modules. This does not assert that each file is a CLI entry point.
• Test modules are .py files in any tests/ directory or named test_* / *_test.py. __init__.py and conftest.py are counted separately as support files, including those outside tests. Support files may contain executable imports or fixtures and need review too.
• Other files include test fixtures and data. None has a tracked executable bit or shebang; discovery aborts if such a non-Python candidate is found, requiring manual classification. Consequently the three skills with no Python helpers have no detected non-Python script executable either. This does not audit executable snippets in SKILL.md.
• Read coverage waivers from config/ci-test-policy.yaml at the assessed revision. Missing or malformed policy fails discovery. “No skill waiver” means no per-skill waiver entry; it is not a measured coverage result. Waiver presence does not itself block an axis pass, nor does meeting a coverage floor prove the axis criteria.

The table lists helper paths relative to skills/<id>/scripts/. Test/support counts refer to that same tree; the reproduction command enumerates their exact membership. All 59 helper-bearing rows have the same outstanding semantic review task: map every helper and relevant support module to happy, boundary, and error fixtures, demonstrate applicable fail-closed behavior and schema conformance, then run and record the exact test command, exit status, revision and date. For outputs without a schema requirement, record why it is inapplicable rather than inventing schema evidence.

Prioritize core risk/state skills (drawdown-circuit-breaker, futures-position-sizer, position-sizer, trader-memory-core) for that review. Then address waiver-bearing skills with #293. Coordinate FMP client evidence with PR #556: its generated clients differ from this snapshot, so evidence from this revision must not be carried forward automatically. Do not infer coverage gaps from helper/test filename mismatches or module-count ratios.

Complete production inventory

Production skill	Python helpers (relative to scripts/)	Test modules / support	Other files	Waiver floor / target / expiry
backtest-expert (not_verified)	evaluate_backtest.py	1 / 1	None	No skill waiver
breadth-chart-analyst (not_verified)	detect_breadth_values.py detect_uptrend_ratio.py extract_chart_right_edge.py fetch_breadth_csv.py	2 / 2	None	25 / 70 / 2027-01-31
breakout-trade-planner (not_verified)	_market_calendar.py order_builder.py plan_breakout_trades.py risk_calculator.py	4 / 1	None	No skill waiver
canslim-screener (not_verified)	calculators/earnings_calculator.py calculators/growth_calculator.py calculators/institutional_calculator.py calculators/leadership_calculator.py calculators/market_calculator.py calculators/new_highs_calculator.py calculators/supply_demand_calculator.py check_institutional_endpoint.py finviz_stock_client.py fmp_client.py report_generator.py scorer.py screen_canslim.py	5 / 1	None	48 / 70 / 2027-01-31
contrarian-setup-gate (not_verified)	gate_logic.py run_contrarian_setup_gate.py	2 / 1	None	No skill waiver
cot-contrarian-detector (not_verified)	cot_index.py fmp_client.py screen_cot_crowding.py	3 / 0	None	No skill waiver
data-quality-checker (not_verified)	check_data_quality.py	1 / 1	None	No skill waiver
dividend-growth-pullback-screener (not_verified)	screen_dividend_growth_rsi.py	2 / 0	None	39 / 70 / 2027-01-31
downtrend-duration-analyzer (not_verified)	analyze_downtrends.py generate_histogram_html.py	3 / 1	None	46 / 70 / 2027-01-31
drawdown-circuit-breaker (not_verified)	_market_calendar.py check_circuit_breaker.py	2 / 1	None	No skill waiver
dual-axis-skill-reviewer (not_verified)	run_dual_axis_review.py	1 / 1	None	No skill waiver
earnings-calendar (not_verified)	_fmp_compat.py fetch_earnings_fmp.py fmp_client.py generate_report.py	3 / 1	None	No skill waiver
earnings-trade-analyzer (not_verified)	_fmp_compat.py _market_calendar.py analyze_earnings_trades.py calculators/gap_size_calculator.py calculators/ma200_calculator.py calculators/ma50_calculator.py calculators/pre_earnings_trend_calculator.py calculators/volume_trend_calculator.py fmp_client.py report_generator.py scorer.py	5 / 2	None	No skill waiver
economic-calendar-fetcher (not_verified)	fmp_client.py get_economic_calendar.py	2 / 0	None	No skill waiver
edge-candidate-agent (not_verified)	auto_detect_candidates.py candidate_contract.py export_candidate.py validate_candidate.py	4 / 1	None	40 / 70 / 2027-01-31
edge-concept-synthesizer (not_verified)	synthesize_edge_concepts.py	1 / 1	None	No skill waiver
edge-hint-extractor (not_verified)	build_hints.py	1 / 1	None	No skill waiver
edge-pipeline-orchestrator (not_verified)	orchestrate_edge_pipeline.py	1 / 1	None	No skill waiver
edge-signal-aggregator (not_verified)	aggregate_signals.py	1 / 1	None	No skill waiver
edge-strategy-designer (not_verified)	design_strategy_drafts.py	1 / 1	None	47 / 70 / 2027-01-31
edge-strategy-reviewer (not_verified)	review_strategy_drafts.py	1 / 1	None	No skill waiver
exposure-coach (not_verified)	calculate_exposure.py	1 / 1	None	No skill waiver
finviz-screener (not_verified)	open_finviz_screener.py	1 / 1	None	No skill waiver
ftd-detector (not_verified)	fmp_client.py ftd_detector.py post_ftd_monitor.py rally_tracker.py report_generator.py	6 / 1	None	No skill waiver
futures-position-sizer (not_verified)	futures_position_sizer.py futures_sizing.py	2 / 1	None	No skill waiver
ibd-distribution-day-monitor (not_verified)	_fmp_compat.py data_loader.py distribution_day_tracker.py exposure_policy.py fmp_client.py history_utils.py ibd_monitor.py math_utils.py models.py report_generator.py risk_classifier.py	10 / 1	None	44 / 70 / 2027-01-31
institutional-flow-tracker (not_verified)	analyze_single_stock.py data_quality.py fmp_client.py track_institution_portfolio.py track_institutional_flow.py	6 / 1	None	No skill waiver
kanchi-dividend-review-monitor (not_verified)	build_review_queue.py	1 / 1	None	No skill waiver
kanchi-dividend-sop (not_verified)	build_entry_signals.py build_sop_plan.py dividend_basis.py event_scanner.py payout_safety.py thresholds.py verdict.py	11 / 1	None	No skill waiver
kanchi-dividend-us-tax-accounting (not_verified)	build_tax_planning_sheet.py	1 / 1	None	No skill waiver
macro-regime-detector (not_verified)	calculators/concentration_calculator.py calculators/credit_conditions_calculator.py calculators/equity_bond_calculator.py calculators/sector_rotation_calculator.py calculators/size_factor_calculator.py calculators/utils.py calculators/yield_curve_calculator.py fmp_client.py macro_regime_detector.py report_generator.py scorer.py	13 / 2	None	No skill waiver
market-breadth-analyzer (not_verified)	calculators/bearish_signal_calculator.py calculators/cycle_calculator.py calculators/divergence_calculator.py calculators/historical_context_calculator.py calculators/ma_crossover_calculator.py calculators/trend_level_calculator.py csv_client.py history_tracker.py market_breadth_analyzer.py report_generator.py scorer.py	10 / 2	None	No skill waiver
market-environment-analysis (not_verified)	_market_calendar.py market_utils.py	2 / 1	None	No skill waiver
market-news-analyst (not_applicable)	None	1 / 0	None	No skill waiver
market-top-detector (not_verified)	_market_calendar.py breadth_csv_client.py calculators/breadth_calculator.py calculators/defensive_rotation_calculator.py calculators/distribution_day_calculator.py calculators/index_technical_calculator.py calculators/leading_stock_calculator.py calculators/math_utils.py calculators/sentiment_calculator.py fmp_client.py historical_comparator.py market_top_detector.py report_generator.py scenario_engine.py scorer.py utils.py	18 / 2	None	No skill waiver
news-reaction-failure-analyzer (not_verified)	analyze_news_reaction.py reaction_math.py	2 / 0	None	No skill waiver
options-strategy-advisor (not_verified)	black_scholes.py	2 / 1	None	No skill waiver
pair-trade-screener (not_verified)	analyze_spread.py find_pairs.py statsmodels_support.py	2 / 1	None	56 / 70 / 2027-01-31
parabolic-short-trade-planner (not_verified)	_fmp_compat.py _market_calendar.py adapters/alpaca_inventory_adapter.py adapters/alpaca_market_data_adapter.py adapters/fixture_market_data_adapter.py adapters/market_data_adapter.py bar_normalizer.py broker_short_inventory_adapter.py calculators/acceleration_calculator.py calculators/atr_calculator.py calculators/liquidity_metrics_calculator.py calculators/ma_extension_calculator.py calculators/parabolic_score_calculator.py calculators/range_expansion_calculator.py check_live_apis.py fmp_client.py generate_pre_market_plan.py intraday_evaluators/first_red_evaluator.py intraday_evaluators/orl_evaluator.py intraday_evaluators/vwap_fail_evaluator.py intraday_size_resolver.py intraday_state_machine.py intraday_state_store.py invalidation_rules.py manual_reasons.py market_clock.py math_helpers.py monitor_intraday_trigger.py parabolic_report_generator.py parabolic_scorer.py plan_builders/first_red_plan_builder.py plan_builders/orl_plan_builder.py plan_builders/vwap_fail_plan_builder.py screen_parabolic.py size_recipe_builder.py ssr_state_tracker.py state_caps.py vwap.py	34 / 5	tests/fixtures/dry_run_minimal.json tests/fixtures/intraday_bars/first_red_clean.json tests/fixtures/intraday_bars/first_red_invalidated.json tests/fixtures/intraday_bars/first_red_same_bar_invalidation_wins.json tests/fixtures/intraday_bars/no_bars_yet.json tests/fixtures/intraday_bars/orl_clean_break.json tests/fixtures/intraday_bars/orl_invalidated.json tests/fixtures/intraday_bars/orl_low_volume_no_fire.json tests/fixtures/intraday_bars/vwap_full_fsm.json tests/fixtures/intraday_bars/vwap_invalidated_after_trigger.json tests/fixtures/phase2_plan_smoke.json	No skill waiver
pead-screener (not_verified)	_fmp_compat.py calculators/breakout_calculator.py calculators/liquidity_calculator.py calculators/risk_reward_calculator.py calculators/weekly_candle_calculator.py fmp_client.py report_generator.py scorer.py screen_pead.py	2 / 2	None	No skill waiver
portfolio-manager (not_verified)	check_alpaca_connection.py reconcile_positions.py	2 / 1	None	No skill waiver
position-sizer (not_verified)	position_sizer.py	1 / 1	None	No skill waiver
pre-trade-discipline-gate (not_verified)	check_pre_trade_discipline.py	1 / 1	None	No skill waiver
scenario-analyzer (not_applicable)	None	1 / 0	None	No skill waiver
sector-analyst (not_verified)	analyze_sector_rotation.py	2 / 1	None	No skill waiver
signal-postmortem (not_verified)	postmortem_analyzer.py postmortem_recorder.py	3 / 1	None	No skill waiver
skill-designer (not_verified)	build_design_prompt.py	1 / 1	None	47 / 70 / 2027-01-31
skill-idea-miner (not_verified)	mine_session_logs.py score_ideas.py	3 / 2	None	No skill waiver
skill-integration-tester (not_verified)	validate_workflows.py	1 / 1	None	No skill waiver
stanley-druckenmiller-investment (not_verified)	allocation_engine.py report_generator.py report_loader.py scorer.py strategy_synthesizer.py	5 / 1	None	No skill waiver
strategy-pivot-designer (not_verified)	detect_stagnation.py generate_pivots.py	4 / 1	None	No skill waiver
technical-analyst (not_verified)	check_weekly_price_action.py weekly_price_action.py	3 / 0	None	No skill waiver
theme-detector (not_verified)	_market_calendar.py calculators/heat_calculator.py calculators/industry_ranker.py calculators/lifecycle_calculator.py calculators/theme_classifier.py calculators/theme_discoverer.py config_loader.py default_theme_config.py etf_scanner.py finviz_performance_client.py leadership.py report_generator.py representative_stock_selector.py scorer.py theme_detector.py theme_history.py theme_matcher.py uptrend_client.py	19 / 2	tests/README.md themes.yaml	No skill waiver
trade-hypothesis-ideator (not_verified)	pipeline/evidence_extractor.py pipeline/format_output.py pipeline/normalize.py pipeline/ranking.py pipeline/strategy_exporter.py run_hypothesis_ideator.py	6 / 2	None	No skill waiver
trader-memory-core (not_verified)	fmp_price_adapter.py thesis_ingest.py thesis_review.py thesis_store.py trader_memory_cli.py	6 / 1	None	No skill waiver
trading-skills-navigator (not_verified)	build_snapshot.py intent_benchmark.py recommend.py	2 / 1	None	No skill waiver
uptrend-analyzer (not_verified)	calculators/historical_context_calculator.py calculators/market_breadth_calculator.py calculators/momentum_calculator.py calculators/sector_participation_calculator.py calculators/sector_rotation_calculator.py data_fetcher.py report_generator.py scorer.py uptrend_analyzer.py	11 / 2	None	No skill waiver
us-market-bubble-detector (not_verified)	bubble_scorer.py	1 / 0	None	No skill waiver
us-stock-analysis (not_applicable)	None	1 / 0	None	No skill waiver
value-dividend-screener (not_verified)	screen_dividend_stocks.py	2 / 1	None	34 / 70 / 2027-01-31
vcp-screener (not_verified)	_fmp_compat.py calculators/execution_state.py calculators/forward_outcome.py calculators/pattern_classifier.py calculators/pivot_proximity_calculator.py calculators/relative_strength_calculator.py calculators/trend_template_calculator.py calculators/vcp_pattern_calculator.py calculators/volume_pattern_calculator.py fmp_client.py historical_report.py historical_scanner.py report_generator.py scorer.py screen_vcp.py	4 / 2	None	No skill waiver
weekly-performance-digest (not_verified)	generate_weekly_digest.py	1 / 1	None	No skill waiver

Production skills: 62. Totals: {‘helpers’: 267, ‘not_applicable’: 3, ‘not_verified’: 59, ‘other’: 13, ‘support’: 68, ‘tests’: 251, ‘with_helpers’: 59, ‘without_helpers’: 3}.

Reproduce the inventory

Run from a checkout containing the assessed commit, with Python 3.9 and PyYAML installed. The command is read-only and emits the table and totals above. It pins all inputs to the assessed revision, even when run from another branch. To inspect exact test/support paths, inspect the groups lists in the command; the classification is explicit.

python3.9 - <<'PY'
import subprocess
from collections import Counter
from pathlib import Path

import yaml

revision = "34af7f8226647c722899a5837eff7e887cf7fbe6"


def read(path):
    return subprocess.check_output(["git", "show", f"{revision}:{path}"])


skills = yaml.safe_load(read("skills-index.yaml"))["skills"]
production = sorted(
    (skill for skill in skills if skill["status"] == "production"),
    key=lambda skill: skill["id"],
)
waivers = yaml.safe_load(read("config/ci-test-policy.yaml"))["coverage"]["waivers"]
files = subprocess.check_output(["git", "ls-tree", "-r", "-z", revision]).decode().split("\0")
tracked = {}
for entry in filter(None, files):
    metadata, path = entry.split("\t", 1)
    mode, kind, _ = metadata.split()
    if kind == "blob":
        tracked[path] = mode

totals = Counter()
print(
    "| Production skill | Python helpers (relative to `scripts/`) | Test modules / support | Other files | Waiver floor / target / expiry |"
)
print("|---|---|---|---|---|")
for skill in production:
    prefix = f"skills/{skill['id']}/scripts/"
    groups = {key: [] for key in ("helpers", "tests", "support", "other")}
    for path in sorted(tracked):
        if not path.startswith(prefix):
            continue
        relative = path[len(prefix) :]
        item = Path(relative)
        is_test = (
            "tests" in item.parts
            or item.name.startswith("test_")
            or item.name.endswith("_test.py")
            or item.name == "conftest.py"
        )
        if item.name in {"__init__.py", "conftest.py"}:
            group = "support"
        elif item.suffix == ".py":
            group = "tests" if is_test else "helpers"
        else:
            group = "other"
            first_line = read(path).split(b"\n", 1)[0]
            if tracked[path] == "100755" or first_line.startswith(b"#!"):
                raise RuntimeError(f"Review non-Python executable before classifying: {path}")
        groups[group].append(relative)
    totals.update({key: len(value) for key, value in groups.items()})
    totals["with_helpers" if groups["helpers"] else "without_helpers"] += 1
    declaration = skill["verification"]["unit_tests"]
    totals[declaration] += 1
    helper_text = "<br>".join(f"`{path}`" for path in groups["helpers"]) or "None"
    other_text = "<br>".join(f"`{path}`" for path in groups["other"]) or "None"
    waiver = waivers.get(skill["id"])
    waiver_text = (
        f"{waiver['floor']} / {waiver['target']} / {waiver['expires_on']}"
        if waiver is not None
        else "No skill waiver"
    )
    print(
        f"| `{skill['id']}` ({declaration}) | {helper_text} | "
        f"{len(groups['tests'])} / {len(groups['support'])} | "
        f"{other_text} | {waiver_text} |"
    )
print()
print(f"Production skills: {len(production)}. Totals: {dict(sorted(totals.items()))}.")
PY

Follow-up evidence record

For each skill, retain helper/support path, test node IDs, exercised behavior, output schema (or applicability rationale), command and exit status, reviewed revision, reviewer and date. Use the production criteria before any passed override; update the baseline and index together only when every requirement is established. Issue #496 remains open for the semantic audit, security review, and instruction contracts.