chore: snapshot local worktree changes by PatrickFanella · Pull Request #631 · PatrickFanella/augr

PatrickFanella · 2026-04-22T17:38:21Z

Summary

snapshots all current local changes from the working tree into a reviewable branch
includes Alpaca reconciliation/runtime/admin endpoint work, discovery changes, tests, migrations, and local plan/task artifacts
intended as a catch-all PR for the current local state, including unrelated work

Notes

this PR intentionally includes unrelated local changes together as requested
branch was created from the current local HEAD and contains the full staged worktree snapshot

Validation

not re-run as a single cohesive PR validation pass before opening; this is a snapshot PR

Copilot

Pull request overview

This snapshot PR bundles several unrelated worktree changes across the web UI, API/runtime wiring, automation (Alpaca reconciliation), discovery strategy deployment deduping, and assorted tests/migrations.

Changes:

Adds Alpaca reconciliation as an automation job + admin API/CLI endpoints, plus metrics and extensive unit tests.
Introduces a migration to deduplicate discovery-generated paper strategies and enforces uniqueness via a partial unique index; bumps required schema version.
Updates the web dashboard/activity UX (historical events + recent runs), plus various robustness tweaks and new component tests.

Reviewed changes

Copilot reviewed 53 out of 53 changed files in this pull request and generated 6 comments.

Show a summary per file

File	Description
web/src/pages/strategies-page.tsx	Removes strategy list filter UI/state; simplifies query to fetch all strategies.
web/src/pages/pipeline-run-page.tsx	Adds safer guards for `phase_timings` and `config_snapshot` rendering.
web/src/pages/order-detail-page.tsx	Refactors “options info” rendering to use typed access helpers instead of repeated casts.
web/src/pages/dashboard-page.tsx	Adds the new “Recent runs” widget to the dashboard layout.
web/src/pages/dashboard-page.test.tsx	Updates dashboard test to stub runs/events endpoints and assert recent runs + activity content.
web/src/components/universe/watchlist-table.tsx	Changes score rendering to rely solely on `score` field (see PR comment re API mismatch).
web/src/components/universe/watchlist-table.test.tsx	Adds basic rendering/navigation tests for the watchlist table.
web/src/components/pipeline/decision-inspector.tsx	Adjusts structured output rendering condition to handle falsy-but-present values.
web/src/components/pipeline/decision-inspector.test.tsx	Adds a unit test covering structured output rendering.
web/src/components/dashboard/recent-runs.tsx	New dashboard widget fetching/displaying recent pipeline runs.
web/src/components/dashboard/recent-runs.test.tsx	Tests recent runs widget (happy path + empty state).
web/src/components/dashboard/activity-feed.tsx	Adds historical event fetch via REST and merges with live websocket feed; improves item rendering.
web/src/components/dashboard/activity-feed.test.tsx	Updates tests to account for REST-seeded events and react-query wrapper.
web/src/components/backtests/backtest-equity-chart.tsx	Makes tooltip formatters resilient to non-number values/labels.
web/src/components/backtests/backtest-equity-chart.test.tsx	Adds tests for non-empty and empty chart states (with recharts mocked).
migrations/discovery_strategy_dedup_migration_test.go	Adds SQL-fragment assertions + integration test for dedup migration behavior.
migrations/000031_discovery_strategy_dedup.up.sql	Deduplicates discovery/options paper strategies and adds partial unique index.
migrations/000031_discovery_strategy_dedup.down.sql	Drops the partial unique index.
internal/repository/postgres/strategy_test.go	Extends test schema to include `options` market_type and adds unique-index rejection test.
internal/repository/postgres/schema_version_test.go	Makes “ahead” test relative to `RequiredSchemaVersion`.
internal/repository/postgres/schema_version.go	Bumps `RequiredSchemaVersion` to 31.
internal/metrics/metrics_test.go	Adds test coverage for new Alpaca reconcile metrics counter.
internal/metrics/metrics.go	Adds `AlpacaReconcileRunsTotal` metric and `RecordAlpacaReconcileRun`.
internal/discovery/orchestrator.go	Uses `CreateOrReusePaperStrategy` to avoid duplicate paper strategy creation.
internal/discovery/options/orchestrator.go	Same reuse logic for options discovery deploy path.
internal/discovery/options/generator_test.go	Adds retry/empty-response tests and verifies prompt schema token changes.
internal/discovery/options/generator.go	Treats empty parsed rules as an error; updates schema docs from `operator`→`op`.
internal/discovery/deploy_test.go	Adds tests for create-or-reuse strategy behavior + unique-violation detection.
internal/discovery/deploy.go	Implements `CreateOrReusePaperStrategy` with “list then create” + unique-violation retry.
internal/data/polygon/universe_test.go	Adds a rate-limit spacing test that currently requires real-time waiting (see PR comment).
internal/data/polygon/universe.go	Increases inter-page delay to 12s to respect Polygon free-tier limits.
internal/cli/root.go	Adds `automation alpaca-reconcile` CLI command that hits the admin endpoint.
internal/cli/automation_test.go	Tests the new CLI command against a stub HTTP server.
internal/automation/orchestrator_test.go	Adds tests for job last summary cloning and Alpaca reconcile metrics/summary behavior.
internal/automation/orchestrator.go	Adds `LastSummary` plumbing, cloning helpers, and registers broker reconciliation jobs.
internal/automation/jobs_broker_reconciliation.go	New scheduled job registration + implementation wrapper for Alpaca reconciliation.
internal/automation/alpaca_reconciliation_test.go	Large unit test suite for reconciliation + verification behavior and orchestrator job presence.
internal/automation/alpaca_reconciliation.go	Implements Alpaca reconcile/verify services + audit logging and snapshot mapping.
internal/automation/alpaca_client_adapter_test.go	Adds adapter tests; pagination test currently doesn’t exercise pagination (see PR comment).
internal/automation/alpaca_client_adapter.go	Implements Alpaca API adapter for orders/fills/positions snapshots.
internal/api/server.go	Wires Alpaca admin endpoints and injects reconciler dependency into the server.
internal/api/request_logger_test.go	Adds tests for request logging suppression behavior.
internal/api/middleware.go	Suppresses request logs for probe/high-volume paths.
internal/api/automation_handlers_test.go	Adds test ensuring automation status includes Alpaca last summary.
internal/api/automation_alpaca_handlers_test.go	Tests the new Alpaca admin reconcile endpoint behavior.
internal/api/automation_alpaca_handlers.go	New handlers/interfaces for `/automation/alpaca/verify` and `/automation/alpaca/reconcile`.
internal/agent/runner_test.go	Adds a panic-in-phase test to ensure run is marked failed and emits PipelineError.
internal/agent/runner.go	Adds panic recovery to mark runs failed + persist terminal events; introduces unconditional DEBUG logs (see PR comment).
cmd/tradingagent/runtime_test.go	Updates schema version expectations; adds test ensuring Alpaca reconcile job is wired.
cmd/tradingagent/runtime.go	Wires Alpaca client adapter + reconciler into API deps and automation orchestrator.
.task/checksum/build	New local task checksum artifact.
.hermes/plans/2026-04-22-alpaca-reconciliation.md	Adds local plan doc for Alpaca reconciliation work.
.hermes/plans/2026-04-20_165747-hermes-integration.md	Removes older Hermes integration plan doc.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-04-22T17:44:50Z

+                <Badge variant={scoreBadgeVariant(t.score)}>
+                  {t.score.toFixed(2)}
                </Badge>


/api/v1/universe/watchlist returns universe.TrackedTicker JSON with a watch_score field (see internal/api/universe_handlers.go), not score. With this change, t.score can be undefined at runtime and toFixed() will throw. Please restore compatibility (e.g., read watch_score when score is absent) or update the API to return score consistently before relying on it here.

Copilot · 2026-04-22T17:44:50Z

+func TestListActiveTickersRespectsFreeTierRateLimit(t *testing.T) {
+	var firstRequestAt time.Time
+	var serverURL string
+
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Header().Set("Content-Type", "application/json")
+
+		switch r.URL.Query().Get("cursor") {
+		case "":
+			firstRequestAt = time.Now()
+			_, _ = fmt.Fprintf(w, `{"results":[{"ticker":"AAA","name":"Alpha","primary_exchange":"XNAS","type":"CS","active":true}],"next_url":"%s/v3/reference/tickers?cursor=page-2"}`,
+				serverURL,
+			)
+		case "page-2":
+			if time.Since(firstRequestAt) < 11*time.Second {
+				w.WriteHeader(http.StatusTooManyRequests)
+				_, _ = w.Write([]byte(`{"status":"ERROR","request_id":"req-rate","error":"rate limit exceeded"}`))
+				return
+			}
+			_, _ = w.Write([]byte(`{"results":[{"ticker":"BBB","name":"Beta","primary_exchange":"XNYS","type":"CS","active":true}]}`))


This test enforces a real-time 11s+ delay to validate rate limiting, which will slow the unit test suite significantly and may cause CI timeouts/flakiness. Consider injecting a clock/sleeper into the Polygon client so the production code can wait 12s while tests can advance time instantly (or assert that the client requested the second page only after the injected delay).

Copilot · 2026-04-22T17:44:51Z

+		// Rate limit pause: Polygon free tier allows 5 req/min, so paginated
+		// reference-ticker requests need roughly 12s spacing to avoid 429s.
 		select {
 		case <-ctx.Done():
 			return tickers, ctx.Err()
-		case <-time.After(250 * time.Millisecond):
+		case <-time.After(12 * time.Second):
 		}


The hard-coded 12 * time.Second sleep will make ListActiveTickers extremely slow for multi-page universes (12s per page) and also makes the behavior hard to test. Consider centralizing this behind a configurable rate limiter / injected sleep function (defaulting to 12s for free-tier) and/or switching to a 429-aware backoff so paid-tier or higher limits aren’t unnecessarily throttled.

Copilot · 2026-04-22T17:44:51Z

+func TestAlpacaClientAdapterListFills_PaginatesActivities(t *testing.T) {
+	t.Parallel()
+
+	var requestCount int
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		if r.Method != http.MethodGet {
+			t.Fatalf("request method = %s, want %s", r.Method, http.MethodGet)
+		}
+		if r.URL.Path != "/v2/account/activities/FILL" {
+			t.Fatalf("request path = %s, want %s", r.URL.Path, "/v2/account/activities/FILL")
+		}
+		requestCount++
+		w.Header().Set("Content-Type", "application/json")
+		switch requestCount {
+		case 1:
+			if got := r.URL.Query().Get("page_token"); got != "" {
+				t.Fatalf("first page_token = %q, want empty", got)
+			}
+			_, _ = w.Write([]byte(`[
+				{
+					"activity_type": "FILL",
+					"id": "fill-2",
+					"order_id": "order-1",
+					"symbol": "SNAL",
+					"side": "buy",
+					"qty": "14",
+					"price": "0.92",
+					"transaction_time": "2026-04-15T19:20:04.943982Z",
+					"order_status": "filled"
+				},
+				{
+					"activity_type": "FILL",
+					"id": "fill-1",
+					"order_id": "order-1",
+					"symbol": "SNAL",
+					"side": "buy",
+					"qty": "186",
+					"price": "0.92",
+					"transaction_time": "2026-04-15T19:20:02.66268Z",
+					"order_status": "partially_filled"
+				}
+			]`))
+		case 2:
+			if got := r.URL.Query().Get("page_token"); got != "fill-1" {
+				t.Fatalf("second page_token = %q, want fill-1", got)
+			}
+			_, _ = w.Write([]byte(`[]`))
+		default:
+			t.Fatalf("unexpected request count %d", requestCount)
+		}
+	}))
+	defer server.Close()
+
+	client := alpacaexec.NewClient("test-key", "test-secret", true, slog.New(slog.NewTextHandler(testDiscardWriter{}, nil)))
+	client.SetBaseURL(server.URL)
+
+	adapter := NewAlpacaClientAdapter(client)
+	fills, err := adapter.ListFills(context.Background())
+	if err != nil {
+		t.Fatalf("ListFills() error = %v", err)
+	}
+	if requestCount != 1 {
+		t.Fatalf("requestCount = %d, want 1", requestCount)
+	}


TestAlpacaClientAdapterListFills_PaginatesActivities doesn’t actually exercise pagination: the first response has 2 items (< page_size=100), so ListFills should stop after 1 request. The test name + server handler’s case 2 suggest you intended to validate the page_token follow-up call. Either rename the test to reflect single-page behavior, or make the first response length equal to alpacaActivitiesPageSize so the adapter must request the second page and you can assert page_token usage.

Copilot · 2026-04-22T17:44:51Z

+	return changed
+}
+
+func fillDedupeKey(fill BrokerFillSnapshot) string {


fillDedupeKey ignores the broker-provided ActivityID and dedupes purely on (ticker, side, qty, price, executed_at). If Alpaca emits multiple fills with identical execution fields (common for small partial fills), this will collapse distinct fills into one trade and/or make reconcile/verify non-idempotent. Consider incorporating ActivityID into the dedupe key and persisting it (e.g., a dedicated column / unique index) so repeated reconciliations can deterministically avoid duplicates without risking dropped fills.

Suggested change

func fillDedupeKey(fill BrokerFillSnapshot) string {

func fillDedupeKey(fill BrokerFillSnapshot) string {

if activityID := strings.TrimSpace(fill.ActivityID); activityID != "" {

return strings.Join([]string{"activity", activityID}, "|")

}

Copilot · 2026-04-22T17:44:52Z

 // Run executes one prepared run and returns the canonical result.
-func (r *Runner) Run(ctx context.Context, prepared PreparedRun) (*RunResult, error) {
+func (r *Runner) Run(ctx context.Context, prepared PreparedRun) (result *RunResult, runErr error) {
 	slog.Info("DEBUG: runner.Run entered", slog.String("run_id", prepared.RunID.String()))


There are unconditional slog.Info("DEBUG: ...") statements using the package-level logger. These will ship as noisy production logs and bypass the runner’s configured logger/level. Please remove them or switch to r.logger.Debug(...) (or equivalent) behind a debug flag so normal runs aren’t spammed.

Suggested change

slog.Info("DEBUG: runner.Run entered", slog.String("run_id", prepared.RunID.String()))

PatrickFanella · 2026-04-22T17:52:31Z

@copilot apply changes based on the comments in this thread

PatrickFanella · 2026-04-22T17:53:10Z

@copilot apply changes based on the comments in this thread

PatrickFanella · 2026-04-22T17:56:46Z

@copilot apply changes based on the comments in this thread

PatrickFanella added 2 commits April 22, 2026 00:13

fix: restore dashboard data and clean frontend build failures

c06b7d8

chore: snapshot local worktree changes

0ade2ab

Copilot AI review requested due to automatic review settings April 22, 2026 17:38

Copilot started reviewing on behalf of PatrickFanella April 22, 2026 17:38 View session

Copilot AI reviewed Apr 22, 2026

View reviewed changes

PatrickFanella assigned Copilot Apr 22, 2026

PatrickFanella added 2 commits April 22, 2026 13:30

fix: address PR review feedback

dcb9b0b

Merge origin/main into chore/local-snapshot-20260422-123615

7ac3612

github-actions Bot added the size:xl Extra large effort label Apr 22, 2026

PatrickFanella merged commit 7c6cb01 into main Apr 22, 2026
3 checks passed

PatrickFanella deleted the chore/local-snapshot-20260422-123615 branch April 22, 2026 19:04

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

chore: snapshot local worktree changes#631

chore: snapshot local worktree changes#631
PatrickFanella merged 4 commits into
mainfrom
chore/local-snapshot-20260422-123615

PatrickFanella commented Apr 22, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Apr 22, 2026

Uh oh!

Copilot AI Apr 22, 2026

Uh oh!

Copilot AI Apr 22, 2026

Uh oh!

Copilot AI Apr 22, 2026

Uh oh!

Copilot AI Apr 22, 2026

Uh oh!

Copilot AI Apr 22, 2026

Uh oh!

PatrickFanella commented Apr 22, 2026

Uh oh!

PatrickFanella commented Apr 22, 2026

Uh oh!

PatrickFanella commented Apr 22, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

PatrickFanella commented Apr 22, 2026

Summary

Notes

Validation

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Apr 22, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 22, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 22, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 22, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 22, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 22, 2026

Choose a reason for hiding this comment

Uh oh!

PatrickFanella commented Apr 22, 2026

Uh oh!

PatrickFanella commented Apr 22, 2026

Uh oh!

PatrickFanella commented Apr 22, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants