ghostfolio

Commit Graph

Author	SHA1	Message	Date
Priyanka Punukollu	11618b0b56	fix: disambiguate share-of queries — my share of AAPL routes to portfolio holdings, share price of AAPL routes to market data Made-with: Cursor	1 month ago
Priyanka Punukollu	453770ce00	perf: Haiku for simple queries, Sonnet for complex — portfolio/property/market queries now ~2s, strategy/life-decision ~10s Made-with: Cursor	1 month ago
Priyanka Punukollu	4e0b14c57a	fix: improve stock price query recognition — handle typos (APPL→AAPL), add share-of phrasing, add company name aliases for ticker extraction Made-with: Cursor	1 month ago
Priyanka Punukollu	283c23e9aa	fix: route stock price queries to market_data not portfolio_analysis Made-with: Cursor	1 month ago
Priyanka Punukollu	47e8c34943	feat: UI polish, chat persistence, auth, parallel evals — 60/60 passing - fix: labels vs buttons — clear visual distinction across login, chat, sidebar - feat: chat persistence on reload — auto-resume last session via localStorage - fix: JWT_SECRET_KEY + ADMIN_PASSWORD_HASH configured; load_dotenv(override=True) - fix: pin bcrypt>=3.2,<4.0 to resolve passlib 1.7.4 compatibility - feat: token-based auth support in run_evals.py (EVAL_AUTH_TOKEN env var) - perf: parallel eval runner with asyncio.gather + semaphore (CONCURRENCY=3) - fix: latency check demoted to warning so API variance never causes false negatives - fix: remove 45s per-request timeout override; use client 65s timeout uniformly - feat: state.py — track input_tokens / output_tokens from Anthropic API - feat: eval_results.md + run_golden_sets.py added Eval result: 60/60 (100%) — adversarial 10/10, edge_case 10/10, happy_path 20/20, multi_step 10/10, write 10/10 Made-with: Cursor	1 month ago
Priyanka Punukollu	8a60e4d719	fix: resolve all eval failures — classifier now passes 267/267 tests at 100% - Fix HP007/HP013: add 'drawdown', 'biggest holding', 'top holdings' to performance keyword lists so these queries route to portfolio_analysis - Fix MS005: use word-boundary regex for short city tokens (sf, atx, dfw) to prevent 'sf' substring-matching inside ticker symbols like 'MSFT', which was incorrectly routing to real_estate_snapshot - Fix MS010: route full_report_kws to performance+compliance+activity (was 'compliance' only, missing transaction_query for 'recent activity') - Fix sc-004: add common 'portfolio' typos (portflio, porfolio, etc.) to natural_performance_kws for robustness against misspellings - Fix MS005 (part 2): add 'worth today', 'worth now', 'currently worth' to market_kws so cost-basis-vs-current-price queries trigger both portfolio_analysis and market_data All eval suites now pass: 182/182 pytest, 60/60 run_evals, 25/25 golden sets Made-with: Cursor	1 month ago
Priyanka Punukollu	4a67afead3	debug: add stdout print logs to classify_node for Railway visibility Made-with: Cursor	1 month ago
Priyanka Punukollu	effa4a217c	fix: broaden context_followup detection in classify_node Follow-up questions like 'how does this compare to inflation?', 'is that good?', 'what does this mean?' were NOT being detected as context_followup because the old check only had 8 hardcoded phrases (e.g. 'how does that compare' but not 'how does this compare'). This caused every follow-up to fall through to 'performance', call portfolio_analysis again, and return the same portfolio summary regardless of what the user actually asked. Added a broader _broad_followup_phrases list (35 patterns) covering: - 'how does this/that/it compare' - 'what does this/that mean' - 'is that good/bad/normal/high/low' - 'compared to inflation/the market' - 'why is that', 'break that down', etc. Confirmed with runtime logs: old_phrase_matched=false for the follow-up, broad_phrase_matched=true, tools_used=[] (correct). Made-with: Cursor	1 month ago
Priyanka Punukollu	cb3674314d	fix: route 'total net worth' to property_net_worth not wealth_portfolio_summary Remove 'total net worth' and 'everything i own' from wealth_net_worth_kws so they fall through to property_net_worth_kws which runs property tracker and shows the complete financial picture template. Add 'show my total net worth' and related phrases to property_net_worth_kws. Made-with: Cursor	1 month ago
Priyanka Punukollu	fa52271e5d	feat: improve net worth response formatting with clear financial picture layout Build pre-formatted financial picture template in property_net_worth executor. Shows investment portfolio, real estate equity per property, total net worth, and portfolio breakdown percentages. If no properties, prompts user to add one. Template is passed as tool result so LLM presents it cleanly. Made-with: Cursor	1 month ago
Priyanka Punukollu	edec30c9c3	feat: add conversational property onboarding flow When user says 'add my home' without property details, return a warm structured prompt asking for address, purchase price, current value, mortgage balance, and monthly rent instead of calling add_property with empty/zero values. Detection: message lacks a price pattern (\$[\d,]+, \d+k, \d{5,}). If message already contains a price, proceed directly to add_property. Made-with: Cursor	1 month ago
Priyanka Punukollu	0726b4e5f5	feat: extract user assumptions from strategy queries — appreciation, interval, years, price Add _extract_strategy_params() helper to graph.py that parses appreciation rate, buy interval, total years, home price, rent yield, and annual income from natural language messages. Update life_decision branch to detect strategy queries and route them to simulate_real_estate_strategy() with extracted params instead of calling the general life_decision_advisor. Supports conservative / moderate / optimistic presets. Falls back to life_decision_advisor for non-strategy queries. Made-with: Cursor	1 month ago
Priyanka Punukollu	1879169f34	fix: update classifier prompt and routing map — all 9 tool categories now correctly routed - Expand classify_node keyword lists for relocation_runway, wealth_gap, equity_unlock, and family_planner to cover more natural phrasings - Add realestate_strategy_kws block that routes multi-property strategy queries ("buy a house every N years", "rental portfolio strategy") to the life_decision executor (home_purchase decision type) - Expand real_estate_kws with MLS/market-data terms and add "show me" to _location_intent_kws so city + intent triggers correctly - Update _route_after_classify with explicit routing map comment showing all 9 non-write categories → "tools" (only existing node names used) - Add tool categories block to SYSTEM_PROMPT so LLM knows to use non-portfolio tool results instead of defaulting to portfolio framing Made-with: Cursor	1 month ago
Priyanka Punukollu	c87072965f	fix: add execution handlers for all 5 new feature query types in graph.py Made-with: Cursor	1 month ago
Priyanka Punukollu	591af17507	feat: add relocation runway calculator Made-with: Cursor	1 month ago
Priyanka Punukollu	dc048d5ef0	feat: complete property_tracker CRUD with SQLite — full agent access - Convert property_tracker.py from in-memory dict to SQLite (agent/data/properties.db) - Add update_property(): update current_value, mortgage_balance, or monthly_rent - Add get_total_net_worth(portfolio_value): combines investment portfolio + real estate equity - Add get_properties(): primary function (list_properties() kept as alias for backward compat) - property_store_clear() now runs DELETE FROM properties (tests stay green) - _row_to_dict() adds backward-compat added_at alias (= created_at) for existing tests - Schema: id TEXT, address, property_type, purchase_price, current_value, mortgage_balance, monthly_rent, county_key, is_active, created_at, updated_at - Soft delete: remove_property() sets is_active = 0 (data preserved for audit) - graph.py: add property_update, property_remove routes + get_total_net_worth execution - graph.py: add "sold my house", "refinanced", "everything i own" keyword triggers Made-with: Cursor	1 month ago
Priyanka Punukollu	3072ea5b74	feat: add wealth_bridge tool — down payment power + global job offer COL calculator - New agent/tools/wealth_bridge.py with 3 registered agent tools: * calculate_down_payment_power(): maps portfolio value to 7 Austin markets (or any cities) with can_afford_full/conservative/safe + monthly payment estimates * calculate_job_offer_affordability(): COL-adjusted salary comparison for any two cities worldwide using ACTRIS COL index (Austin) + Teleport API (global) * get_portfolio_real_estate_summary(): reads live Ghostfolio portfolio + runs down payment analysis in one call - graph.py: add wealth_bridge imports + 4 new query_type routes: wealth_down_payment, wealth_job_offer, wealth_global_city, wealth_portfolio_summary - graph.py: add _extract_salary, _extract_offer_city, _extract_current_city helpers - Mortgage formula: 30yr @ 6.95%, 20% down, ×1.25 for PITI Made-with: Cursor	1 month ago
Priyanka Punukollu	e1bdb2fc88	feat(agent): complete showcase — real ACTRIS data, property tracker, 27 UI features Backend: - Replace Austin mock data with real Jan 2026 ACTRIS/Unlock MLS figures covering 7 counties/MSAs (Travis, Williamson, Hays, Bastrop, Caldwell, Austin MSA) - Add property_tracker tool: add/list/remove properties with equity & gain calc - Expand graph.py routing for property tracking intent + 15 new city aliases - Add /health endpoint and improved SSE streaming in main.py - 81 passing pytest evals (test_portfolio, test_property_tracker, test_real_estate) Frontend (chat_ui.html) — 27 new features: - Command palette (Cmd+P) with 29 commands and fuzzy search - Cross-session search across all saved chats in the drawer - User investor profile (risk/focus/horizon) injected as AI context - Rental yield calculator (gross/net yield, cap rate, annual income) - Portfolio donut chart (SVG, click-to-query slices) - Property comparison table, market calendar strip, county alerts - Smart time-based suggestions, offline detection + message queue - PWA manifest + install prompt, scroll-to-bottom button - High contrast mode, reduced motion, keyboard focus rings - Response disclaimer toggle, copy-as-Markdown, batch export - Email digest, session reminders, conversation branching - Swipe-to-archive (mobile), collaborative annotations via URL - Settings menu scrollable (max-height fix for 20+ items) - Fix: broken paddingRight string literal silently killed all event listeners - Fix: extra stray </div> in help panel causing HTML parse error Angular ai-chat component: - Fix: prefer-optional-chain lint error in successRate() - Fix: prettier formatting on chat_ui.html - Add portfolio-chart and real-estate-card sub-components All 81 pytest evals pass. Lint: 0 errors. Prettier: all files formatted. Made-with: Cursor	1 month ago
Priyanka Punukollu	c639ecdae3	feat(real-estate): expand keyword triggers + bedroom/price filter - classify_node: add homes, houses, bedroom, 3 bed, 2br, under $, rent estimate, for sale, find homes, open house, property search. - classify_node: add real_estate_search sub-triggers for bedroom/ price queries; add real_estate_detail route for listing ID patterns. - tools_node: wire real_estate_detail branch → get_listing_details(). - tools_node: pass min_beds/max_price extracted from query to search_listings(). - graph.py: add _extract_search_filters() helper (beds + price parsing). - graph.py: import get_listing_details from real_estate module. - format_node: handle nested {code, message} error dicts from tools. - format_node: inject _re_context guard when query_type starts with real_estate to prevent Claude from claiming it lacks RE data. Made-with: Cursor	1 month ago
Priyanka Punukollu	fac75947e9	feat: connect real-estate tool to LangGraph agent - graph.py: import real_estate functions (get_neighborhood_snapshot, search_listings, compare_neighborhoods, is_real_estate_enabled). - classify_node: detect real estate keywords → route to real_estate_snapshot \| real_estate_search \| real_estate_compare. Detection is fully guarded by is_real_estate_enabled() — when flag is off, this block is never entered and existing routing is unchanged. - tools_node: 3 new elif branches for real estate query types. All append-only; no existing branches modified. - Added _extract_real_estate_location() and _extract_two_locations() helpers (new functions, no changes to existing helpers). Made-with: Cursor	1 month ago
Priyanka Punukollu	3aa078db3b	fix: achieve 25/25 evals — robust criteria + health check routing - eval runner: add retry logic (2 attempts) for transient connection drops - gs-001: accept 'percent' as well as '%' (LLM formatting variance) - gs-002: use must_contain_one_of for ticker/company name variance - gs-008/sc-014: fix expected_tools for conditionally-triggered compliance - graph.py: route 'health check'/'full report' queries to compliance path so compliance_check always runs for full portfolio report requests Co-authored-by: Cursor <cursoragent@cursor.com>	1 month ago
Priyanka Punukollu	b7619dd562	fix: reduce citation spam — cite once per sentence not after every number Source tags [tool_result_id] were appearing after every individual figure, making responses unreadable. Rules 1 and 10 in SYSTEM_PROMPT and the format_node user prompt now enforce one citation per sentence placed at the end, not inline after each value. Co-authored-by: Cursor <cursoragent@cursor.com>	1 month ago
Priyanka Punukollu	5c6bda0ea4	feat(agent): add login page, live thinking steps, and UI polish - Login page (login.html) with email/password auth, error states, demo hint - /auth/login FastAPI endpoint with credential validation - /chat/steps SSE endpoint streaming real-time LangGraph node events - /me endpoint for user profile lookup - chat_ui.html: auth guard, sign-out, localStorage persistence, category quick prompts, live thinking panel, tool badges, confidence bar, verification badge, copy button, retry button, latency tracker, session summary toast, /tools command, message timestamps Co-authored-by: Cursor <cursoragent@cursor.com>	1 month ago
Priyanka Punukollu	f9672042d3	Revert "feat: AI portfolio agent — LangGraph, 6 tools, golden sets, 60/60 evals" This reverts commit `a62faae8dd`.	1 month ago
Priyanka Punukollu	a62faae8dd	feat: AI portfolio agent — LangGraph, 6 tools, golden sets, 60/60 evals Co-authored-by: Cursor <cursoragent@cursor.com>	1 month ago

25 Commits (11618b0b566059f60a37054bf6c234899bbeb7fa)