Commit Graph

4 Commits

Author SHA1 Message Date
Félix Malfait 1088f7bbab feat: add lingui/no-unlocalized-strings ESLint rule and fix translations (#16610)
## Summary
This PR adds the `lingui/no-unlocalized-strings` ESLint rule to detect
untranslated strings and fixes translation issues across multiple
components.

## Changes

### ESLint Configuration (`eslint.config.react.mjs`)
- Added comprehensive `ignore` patterns for non-translatable strings
(CSS values, HTML attributes, technical identifiers)
- Added `ignoreNames` for props that don't need translation (className,
data-*, aria-*, etc.)
- Added `ignoreFunctions` for console methods, URL APIs, and other
non-user-facing functions
- Disabled rule for debug files, storybook, and test files

### Components Fixed (~19 files)
- Object record components (field inputs, pickers, merge dialogs)
- Settings components (accounts, admin panel)
- Serverless function components
- Record table and title cell components

## Status
🚧 **Work in Progress** - ~124 files remaining to fix

This PR is being submitted as draft to allow progressive fixing of
remaining translation issues.

## Testing
- Run `npx eslint "src/**/*.tsx"` in `packages/twenty-front` to check
remaining issues
2025-12-17 22:08:33 +01:00
Félix Malfait e7ebf51e50 Replace agent handoff system with planning-based router (#16003)
## Overview

This PR replaces the dynamic agent handoff system with a more
predictable planning-based router that decides upfront how to handle
multi-agent coordination.

## Major Changes

### 🔄 Architecture Shift: Handoffs → Planning

**Removed:**
- `AgentHandoffEntity` and handoff tracking system
- `AgentHandoffService` and `AgentHandoffExecutorService`
- Dynamic agent-to-agent transfers during execution
- Handoff tool generation and description templates

**Added:**
- `AiRouterService` with two strategies: `simple` (single agent) and
`planned` (multi-agent)
- `AgentPlanExecutorService` for executing multi-step plans
- Plan validation (cycle detection, dependency resolution)
- `UnifiedRouterResult` type with discriminated union

### 🤖 New Standard Agents

Added two new specialized agents:
- **Researcher Agent**: Web search, fact-finding, competitive
intelligence
- **Code Agent**: TypeScript function generation for serverless
workflows

### 🏗️ Router Refactoring (Latest)

Split router responsibilities into focused services:
- `AiRouterStrategyDeciderService`: Decides simple vs planned strategy
- `AiRouterPlanGeneratorService`: Generates and validates execution
plans
- `AiRouterService`: Coordinates between services (reduced from 426→275
lines)

### ⚙️ Configuration Improvements

- Added `outputStrategy` to agent definitions (`direct` vs `synthesize`)
- Removed hardcoded special cases for workflow-builder
- Added `plannerModel` field to workspace entity
- Increased `MAX_STEPS` from 10 to 25 for complex workflows

### 📝 Agent Prompt Refinements

Significantly simplified prompts for better clarity:
- Workflow Builder: 51→36 lines
- Helper: 49→28 lines
- Data Manipulator: Enhanced with sorting guidance

### 🔍 Enhanced Debugging

- Plan reasoning and step count in data message parts
- Router debug info with token usage tracking
- Better logging throughout execution pipeline

## Benefits

1. **Simpler Mental Model**: Router decides upfront vs dynamic transfers
2. **Better Predictability**: Users see the plan before execution
3. **Cleaner Architecture**: SRP with focused services
4. **Configuration Over Code**: Agent behavior via config, not hardcoded
logic
5. **Plan Validation**: Catches invalid dependencies and cycles

## Migration Notes

- Database migration removes `agentHandoff` table
- Adds `plannerModel` column to workspace table
- No API breaking changes (agent endpoints unchanged)

## Testing

- Integration tests updated to remove handoff dependencies
- Agent tool test utilities simplified
- Plan validation covered by new logic

## Next Steps (Future PRs)

- Parallel execution of independent plan steps
- Dynamic re-planning based on results
- Plan caching for common routing patterns
- Error recovery strategies in plan executor
2025-11-25 12:10:14 +01:00
Félix Malfait cd5d0f9eac Introduce agent hints to reduce context bloat (#15763)
Introducing a new pattern that should reduce token consumption by 90%
for the most common use-cases
2025-11-13 16:15:01 +01:00
Félix Malfait f9c61833ec Add debug info in AI chat (#15758)
## 🐛 Critical Bug Fix

### Cost Calculation Error (1000x undercharge)
- **Fixed**: Cost conversion utility was calculating credits at 1/1000th
of actual value
- **Before**: `cents * 10` 
- **After**: `(cents / 100) * DOLLAR_TO_CREDIT_MULTIPLIER` 
- **Impact**: Users were being undercharged by 1000x
  - Example: 0.75 cents should = 7,500 credits
  - Bug calculated it as 7.5 credits

---

## 🎯 Code Centralization & DRY

### Unified Cost Calculation
- Centralized all cost conversions to use `convertCentsToBillingCredits`
utility
- Refactored 3 different implementations into 1 single source of truth
- Files updated:
  - `ai-billing.service.ts`
  - `agent-streaming.service.ts` (2 usages)

**Before** (multiple implementations):
```typescript
// Wrong implementation
const credits = cents * 10;

// Verbose implementation  
const costInDollars = costInCents / 100;
const creditsUsed = Math.round(costInDollars * DOLLAR_TO_CREDIT_MULTIPLIER);
```

**After** (unified):
```typescript
const creditsUsed = Math.round(convertCentsToBillingCredits(costInCents));
```

---

##  UI Component Refactoring

### RoutingDebugDisplay.tsx
- **Reduced from 118 lines to 34 lines** (71% reduction)
- Extracted `renderTimingRow` helper to eliminate 15 repetitive JSX
blocks
- Added `formatTokenBreakdown` helper for token display logic
- Much easier to add new debug metrics

**Before**: 15 nearly-identical blocks of repetitive JSX  
**After**: Clean, DRY implementation with reusable helpers

---

## 🧹 Code Quality Improvements

### Removed Debug Code
- Removed `console.log` accidentally left in `RoutingStatusDisplay.tsx`

### Cleaned Up Comments (18+ removed)
Removed redundant comments that stated the obvious:
-  "Calculate routing cost if we have token usage"
-  "Send the updated routing status with execution metrics to the
client"
-  "Count tool calls in the response"
-  "AI SDK's LanguageModelUsage uses inputTokens/outputTokens"
- And 14+ more...

Kept meaningful comments:
-  "Timing is optional, ignore errors" (explains catch block)
-  Type definition grouping comments

---

## 📊 Statistics

**Files Modified**: 10
- `convert-cents-to-billing-credits.util.ts` (fixed formula)
- `ai-billing.service.ts` (use centralized utility)
- `agent-streaming.service.ts` (use utility, remove comments)
- `agent-execution.service.ts` (remove comments)
- `ai-router.service.ts` (remove comments)
- `RoutingStatusDisplay.tsx` (remove debug code)
- `RoutingDebugDisplay.tsx` (major refactor) 
- `isDebugModeState.ts` (new file)
- `DataMessagePart.ts` (type extensions)
- `useClientConfig.ts` (debug mode support)

**Impact**:
- Lines removed: ~130 (redundant code + comments)
- Lines added: ~45 (helper functions)
- **Net reduction**: ~85 lines
- **Bug fixes**: 1 critical (1000x cost error)
- **Centralizations**: 3 locations now using shared utility
- **Major refactors**: 1 UI component (71% reduction)

---

##  Verification

-  All linter checks pass
-  All tests pass (`ai-billing.service.spec.ts` verified)
-  No `any` types in affected code
-  No TODO/FIXME markers

---

## 🎯 Principles Applied

1.  **Fix Root Causes, Not Symptoms** - Fixed utility function, then
used it everywhere
2.  **DRY (Don't Repeat Yourself)** - Centralized cost calculation and
UI rendering
3.  **Single Source of Truth** - One place for cost conversion formula
4.  **Code as Documentation** - Removed comments that repeated what
code says
5.  **Composability** - Created reusable helper functions
6.  **Type Safety** - Maintained strict typing throughout
2025-11-13 13:34:46 +01:00