kai/smart-support
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
24 Commits 1 Branch 0 Tags
be5c84bcffa0bf4bc3eb85f58ab10679f7421b12
Commit Graph

24 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Yaojia Wang
be5c84bcff docs: reconcile README and docs with actual codebase 
README:
- Remove duplicated agent config, safety, security sections (covered by docs)
- Add ux_design_system.md and safety.py to project structure and doc links
- Convert doc links to descriptive table

agent-config-guide.md:
- Replace fictional agents/tools with real ones from agents.yaml
- Remove nonexistent 'admin' permission level (only read/write)
- Fix template names (e-commerce, saas, fintech)
- List all available built-in tools

openapi-import-guide.md:
- Fix /result -> /classifications endpoint
- Fix POST /approve to show no request body
- Remove nonexistent 'admin' access type
- Update response examples to match actual API

demo-script.md:
- Fix agent names (order_agent -> order_lookup)
- Replace fictional refund scenario with real lookup+cancel flow

ARCHITECTURE.md:
- Fix langgraph-supervisor version (v1.1 -> 0.0.12+)

docker-compose.yml:
- Expose postgres on port 5433 for local dev
 2026-04-06 13:55:45 +02:00
Yaojia Wang
19fc9f3289 test: close coverage gaps and add frontend test infrastructure 
Backend (516 tests, 94% coverage):
- Add azure_openai endpoint/deployment validation tests (config.py -> 100%)
- Add _total_conversations and _avg_turns direct tests (queries.py -> 100%)
- Add transformer edge cases: list content, string checkpoint, invalid JSON,
  malformed message graceful skip (transformer.py -> 93%)
- Add safety combined status_code+error_message interaction tests
- Fix ambiguous 200/422 assertion to strict 422
- Add E2E pagination shape assertions (total, page, per_page, row count)
- Fix ReplayPool mock to respect LIMIT/OFFSET params

Frontend (23 tests, vitest + happy-dom + @testing-library/react):
- Add vitest infrastructure with happy-dom environment
- Add api.ts tests: success, HTTP error, success=false, URL encoding
- Add DashboardPage tests: loading, data, error, empty states
- Add ReplayListPage tests: loading, empty, data, error, status badge classes
- Add ReplayPage tests: loading, steps, empty, error states
 2026-04-06 13:32:10 +02:00
Yaojia Wang
036e12349d refactor: formalize safety rules, extract shared styles, reconcile docs (P2) 
- Add backend/app/safety.py with explicit confirmation policy, multi-intent
  semantics, and MCP error taxonomy with retry classification
- Add 26 unit tests for safety module (confirmation rules, error taxonomy)
- Extract repeated inline styles into shared CSS classes in index.css
  (section-card, stat-label, status-badge, data-table, empty/error-state,
  pagination-bar)
- Refactor DashboardPage, ReplayListPage, ReplayPage to use shared classes
- Update README: add missing API endpoints, document safety/confirmation rules
- Use proper HTML entities for arrow/dash characters to fix encoding glitches
 2026-04-05 23:10:50 +02:00
Yaojia Wang
e0931daece feat: wire frontend pages to live APIs and standardize response contracts (P1) 
- Backend: Add COUNT query and paginated response shape to conversations endpoint
  Returns { conversations: [...], total, page, per_page } instead of flat array
- Frontend: Replace mock data in DashboardPage with fetchAnalytics() API calls
- Frontend: Replace mock data in ReplayListPage with fetchConversations() API calls
- Frontend: Replace mock data in ReplayPage with fetchReplay() API calls
- Add proper loading, empty, and error states to all three pages
- Align ConversationSummary type with actual DB columns (created_at, status)
- Update unit and E2E tests for new paginated conversation response shape
- Add fetchone() to FakeCursor for COUNT query support in E2E tests
 2026-04-05 23:06:00 +02:00
Yaojia Wang
e55ec42ae5 fix: restore green builds and align frontend-backend contracts (P0) 
- Isolate Settings tests from .env and process env leakage
- Fix analytics metadata test to unwrap psycopg Json wrapper
- Remove unused state variables causing frontend build failures
- Fix ReviewPage to use /classifications endpoint instead of nonexistent /result
- Normalize ReviewPage status enums (failed not error) and access_type values
- Align api.ts types with backend response shapes (ReplayPage, AnalyticsData, AgentUsage)
 2026-04-05 23:00:39 +02:00
Yaojia Wang
189a0fad34 feat(ui): implement premium beige design system and ux refinements 2026-04-05 22:35:48 +02:00
Yaojia Wang
d2b4610df9 fix: address code and security review findings for Phase 5 
- Add nginx security headers (X-Frame-Options, X-Content-Type-Options, etc.)
- Fix postgres networking: add to app_network, comment out host port exposure
- Fix rate limit memory leak: add bounded eviction for stale thread entries
- Use immutable update pattern in rate limit check (no .append mutation)
- Extract _VERSION constant to avoid duplicate hardcoded version string
 2026-03-31 21:35:13 +02:00
Yaojia Wang
0e78e5b06b feat: complete phase 5 -- error hardening, frontend, Docker, demo, docs 
Backend:
- ConversationTracker: Protocol + PostgresConversationTracker for lifecycle tracking
- Error handler: ErrorCategory enum, classify_error(), with_retry() exponential backoff
- Wire PostgresAnalyticsRecorder + ConversationTracker into ws_handler
- Rate limiting (10 msg/10s per thread), edge case hardening
- Health endpoint GET /api/health, version 0.5.0
- Demo seed data script + sample OpenAPI spec

Frontend (all new):
- React Router with NavBar (Chat / Replay / Dashboard / Review)
- ReplayListPage + ReplayPage with ReplayTimeline component
- DashboardPage with MetricCard, range selector, zero-state
- ReviewPage for OpenAPI classification review
- ErrorBanner for WebSocket disconnect handling
- API client (api.ts) with typed fetch wrappers

Infrastructure:
- Frontend Dockerfile (multi-stage node -> nginx)
- nginx.conf with SPA routing + API/WS proxy
- docker-compose.yml with frontend service + healthchecks
- .env.example files (root + backend)

Documentation:
- README.md with quick start and architecture
- Agent configuration guide
- OpenAPI import guide
- Deployment guide
- Demo script

48 new tests, 449 total passing, 92.87% coverage
 2026-03-31 21:20:06 +02:00
Yaojia Wang
38644594d2 test: add thread_id validation tests for replay API 
- Test invalid thread_id with spaces returns 400
- Test thread_id with special chars returns 400
- Tighten existing 404 test assertion
 2026-03-31 13:44:04 +02:00
Yaojia Wang
ef6e5ac2be fix: address security findings in Phase 4 analytics and replay 
- Fix CRITICAL: use parameterized INTERVAL arithmetic (%(days)s * INTERVAL '1 day')
  instead of string interpolation inside SQL literal
- Use asyncio.gather() for parallel query execution in get_analytics()
- Add range upper bound (max 365 days) to prevent DoS via full-table scans
- Add thread_id validation (alphanumeric, max 128 chars) in replay API
- Sanitize error messages to not reflect user input
 2026-03-31 13:38:09 +02:00
Yaojia Wang
33db5aeb10 feat: complete phase 4 -- conversation replay API + analytics dashboard 
- Replay models: StepType enum, ReplayStep, ReplayPage frozen dataclasses
- Checkpoint transformer: PostgresSaver JSONB -> structured timeline steps
- Replay API: GET /api/conversations (paginated), GET /api/replay/{thread_id}
- Analytics models: AgentUsage, InterruptStats, AnalyticsResult
- Analytics event recorder: Protocol + PostgresAnalyticsRecorder + NoOp
- Analytics queries: resolution_rate, agent_usage, escalation_rate, cost, interrupts
- Analytics API: GET /api/analytics?range=Xd with envelope response
- DB migration: analytics_events table + conversations column additions
- 74 new tests, 399 total passing, 92.87% coverage
 2026-03-31 13:35:45 +02:00
Yaojia Wang
a2f750269d fix: address critical security and code review findings in Phase 3 
- Wire ImportOrchestrator into review_api start_import via BackgroundTasks
- Sanitize docstrings in generated tool code to prevent code injection
- Add Literal["read", "write"] validation for access_type
- Add regex validation for agent_group
- Validate URL scheme (http/https only) in ImportRequest
- Validate LLM output fields (clamp confidence, validate access_type)
- Use dataclasses.replace instead of manual reconstruction in importer
- Expand SSRF blocked networks (Carrier-Grade NAT, IPv4-mapped IPv6, etc.)
- Make _BLOCKED_NETWORKS immutable tuple
- Use yaml.safe_dump instead of yaml.dump
- Fix _to_snake_case for empty strings and Python keywords
 2026-03-31 00:28:28 +02:00
Yaojia Wang
a54eb224e0 feat: complete phase 3 -- OpenAPI auto-discovery, SSRF protection, tool generation 
- SSRF protection: private IP blocking, DNS rebinding defense, redirect validation
- OpenAPI fetcher with SSRF guard, JSON/YAML auto-detection, 10MB limit
- Structural spec validator (3.0.x/3.1.x)
- Endpoint parser with $ref resolution, auto-generated operation IDs
- Heuristic + LLM endpoint classifier with Protocol interface
- Review API at /api/openapi (import, job status, classification CRUD, approve)
- @tool code generator + Agent YAML generator
- Import orchestrator (fetch -> validate -> parse -> classify pipeline)
- 125 new tests, 322 total passing, 93.23% coverage
 2026-03-31 00:10:44 +02:00
Yaojia Wang
006b4ee5d7 fix: resolve ruff lint errors in Phase 2 code 
- Move intent imports to TYPE_CHECKING block in graph.py (TC001)
- Rename test classes to CapWords convention (N801)
- Fix line length violations across test files (E501)
- Auto-fix import sorting (I001)
 2026-03-30 21:44:47 +02:00
Yaojia Wang
b861ff055f test: add routing integration tests for Phase 2 test requirements 
9 tests covering the complete multi-agent routing flow:
- Single-intent routing to each agent (order_lookup, order_actions, discount, fallback)
- Multi-intent routing hint injection for sequential execution
- Ambiguity detection skips graph and returns clarification
- Low confidence threshold triggers ambiguity
- No-classifier fallback to supervisor prompt routing

Fills Phase 2 test requirement for integration-level routing coverage.
Total: 197 tests, 92.60% coverage.
 2026-03-30 21:41:01 +02:00
Yaojia Wang
512f988dd0 test: add Phase 2 checkpoint acceptance tests 
18 integration tests validating all 7 Phase 2 checkpoint criteria:
1. Order query routes to order_lookup agent
2. Multi-intent classification with routing hint injection
3. Ambiguous message triggers clarification prompt
4. 30-min interrupt TTL auto-cancel with retry prompt
5. Webhook POST escalation with retry on failure
6. E-commerce template loads 4 correctly configured agents
7. Coverage at 92.60% (188 tests total)
 2026-03-30 21:38:25 +02:00
Yaojia Wang
6e7b824b64 test: add integration tests for WebSocket message flow 
17 integration tests covering:
- Happy path: token streaming, tool calls, multi-message sessions
- Interrupt flow: approve and reject paths with manager tracking
- Session TTL: expiration, sliding window reset, interrupt extension
- Validation: invalid JSON, missing fields, size limits
- Interrupt TTL: expired interrupt sends retry prompt

Fills Phase 1 test gap for integration-level WebSocket coverage.
Total: 170 tests, 92.15% coverage.
 2026-03-30 21:24:31 +02:00
Yaojia Wang
1050df780d feat: complete phase 2 -- multi-agent routing, interrupt TTL, escalation, templates 
- Intent classification with LLM structured output (single/multi/ambiguous)
- Discount agent with apply_discount and generate_coupon tools
- Interrupt manager with 30-min TTL auto-expiration and retry prompts
- Webhook escalation module with exponential backoff retry (max 3)
- Three vertical industry templates (e-commerce, SaaS, fintech)
- Template loading in AgentRegistry
- Enhanced supervisor prompt with dynamic agent descriptions
- 153 tests passing, 90.18% coverage
 2026-03-30 21:04:39 +02:00
Yaojia Wang
7c3571b47d chore: update local Claude Code permission settings 
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-03-30 15:14:16 +02:00
Yaojia Wang
98331abbd5 Update doc 2026-03-30 15:09:04 +02:00
Yaojia Wang
47d4a57313 docs: mark phase 1 as completed 2026-03-30 00:54:54 +02:00
Yaojia Wang
33488fd634 feat: complete phase 1 -- core framework with chat loop, agents, and React UI 
Backend:
- FastAPI WebSocket /ws endpoint with streaming via LangGraph astream
- LangGraph Supervisor connecting 3 mock agents (order_lookup, order_actions, fallback)
- YAML Agent Registry with Pydantic validation and immutable configs
- PostgresSaver checkpoint persistence via langgraph-checkpoint-postgres
- Session TTL with 30-min sliding window and interrupt extension
- LLM provider abstraction (Anthropic/OpenAI/Google)
- Token usage + cost tracking callback handler
- Input validation: message size cap, thread_id format, content length
- Security: no hardcoded defaults, startup API key validation, no input reflection

Frontend:
- React 19 + TypeScript + Vite chat UI
- WebSocket hook with reconnect + exponential backoff
- Streaming token display with agent attribution
- Interrupt approval/reject UI for write operations
- Collapsible tool call viewer

Testing:
- 87 unit tests, 87% coverage (exceeds 80% requirement)
- Ruff lint + format clean

Infrastructure:
- Docker Compose (PostgreSQL 16 + backend)
- pyproject.toml with full dependency management
 2026-03-30 00:54:21 +02:00
Yaojia Wang
e4f08576a9 docs: add CLAUDE.md project instructions and architecture docs 
- CLAUDE.md with phase execution workflow and project conventions
- ARCHITECTURE.md with system design, data flow, and component breakdown
- DEVELOPMENT-PLAN.md with detailed 5-phase task breakdown
 2026-03-30 00:26:50 +02:00
Yaojia Wang
f93e8baef1 feat: initial project setup with planning docs 
Smart Support - AI customer service action layer framework.
Includes design doc, CEO plan, eng review, test plan, and README.
 2026-03-29 21:11:36 +02:00