@hienlh/ppm 0.9.79 → 0.9.81

package/.opencode/workflows/sales-workflow.md

@@ -0,0 +1,193 @@
+# Sales Workflow
+
+Lead-to-customer journey automation from generation to expansion.
+
+## Overview
+
+| Field | Value |
+|-------|-------|
+| **Purpose** | Convert leads to customers, expand accounts |
+| **Agents** | attraction-specialist, lead-qualifier, email-wizard, sale-enabler, upsell-maximizer |
+| **Human Checkpoints** | High-value lead handoff, Close deals |
+
+---
+
+## Workflow Stages
+
+### Stage 1: Lead Generation
+
+**Agent:** `attraction-specialist`
+
+**Actions:**
+1. Create lead magnets
+2. SEO content for TOFU
+3. Landing page optimization
+4. Lead capture forms
+
+**Output:** Lead gen assets, captured leads
+
+**Prompt Template:**
+```
+Create lead generation assets for [product/service].
+Target: [audience]
+Include: lead magnet concept, landing page copy, SEO keywords
+```
+
+---
+
+### Stage 2: Lead Qualification
+
+**Agent:** `lead-qualifier`
+
+**Actions:**
+1. Score leads (demographic + behavioral)
+2. Segment by intent
+3. Identify sales-ready leads
+4. Flag high-value prospects
+
+**Output:** Qualified lead list, scoring report
+
+**Prompt Template:**
+```
+Qualify leads from [source].
+Scoring criteria: [criteria]
+Segment into: cold, warm, hot
+Flag high-value prospects for human review.
+```
+
+---
+
+### Stage 3: Nurture Sequence
+
+**Agent:** `email-wizard`
+
+**Actions:**
+1. Design drip campaigns
+2. Personalize content
+3. A/B test sequences
+4. Optimize timing
+
+**Output:** Email sequences in `content/emails/`
+
+**Prompt Template:**
+```
+Create nurture sequence for [segment].
+Journey: awareness → consideration → decision
+Length: [X] emails over [Y] days
+Include: subject lines, body, CTAs
+```
+
+---
+
+### Stage 4: Sales Enablement
+
+**Agent:** `sale-enabler`
+
+**Actions:**
+1. Create pitch decks
+2. Build objection handlers
+3. Generate case studies
+4. Prepare proposals
+
+**Output:** Sales collateral in `content/sales/`
+
+**Human Checkpoint:** High-value deal support
+
+**Prompt Template:**
+```
+Create sales materials for [prospect/segment].
+Include: pitch deck, objection responses, relevant case study.
+Customize for: [pain points, industry]
+```
+
+---
+
+### Stage 5: Upsell & Cross-sell
+
+**Agent:** `upsell-maximizer`
+
+**Actions:**
+1. Identify expansion opportunities
+2. Product recommendations
+3. Upgrade campaigns
+4. Retention offers
+
+**Output:** Expansion campaign assets
+
+**Prompt Template:**
+```
+Create upsell strategy for [customer segment].
+Current product: [product]
+Expansion targets: [products/features]
+Include: messaging, offers, timing
+```
+
+---
+
+## Agent Orchestration
+
+```
+[Generation] → [Qualification] → [Nurture] → [Enablement] → [Expansion]
+      │              │              │             │              │
+attraction-     lead-         email-       sale-         upsell-
+specialist     qualifier       wizard      enabler       maximizer
+```
+
+---
+
+## Lead Scoring Model
+
+| Signal | Points | Weight |
+|--------|--------|--------|
+| Email open | +5 | Low |
+| Link click | +10 | Medium |
+| Content download | +20 | High |
+| Pricing page visit | +30 | Very High |
+| Demo request | +50 | Critical |
+| Company size match | +15 | Medium |
+| Industry fit | +15 | Medium |
+
+**Thresholds:**
+- Cold: 0-30 points
+- Warm: 31-60 points
+- Hot: 61+ points
+- Sales-ready: 80+ points
+
+---
+
+## Integration Points
+
+### Hooks Triggered
+- `campaign-tracking` → Lead captured, stage transition
+- `brand-guidelines-reminder` → Before content creation
+
+### Data Flow
+- Lead capture → Qualification
+- Scoring → Nurture segment assignment
+- Engagement → Score updates
+- Sales-ready → Human handoff
+
+---
+
+## Usage Example
+
+```
+/funnel design "SaaS Trial Conversion"
+
+# Triggers workflow:
+1. attraction-specialist: Lead magnet, landing page
+2. lead-qualifier: Setup scoring, segments
+3. email-wizard: Nurture sequences by segment
+4. sale-enabler: Demo support materials
+5. upsell-maximizer: Trial → Paid conversion offers
+```
+
+---
+
+## Success Metrics
+
+- Lead volume (MQLs generated)
+- Qualification accuracy (MQL → SQL rate)
+- Nurture engagement (open rate, click rate)
+- Conversion rate (lead → customer)
+- Expansion revenue (upsell success)

package/.opencode/workflows/seo-workflow.md

@@ -0,0 +1,248 @@
+# SEO Workflow
+
+SEO optimization process from audit to monitoring.
+
+## Overview
+
+| Field | Value |
+|-------|-------|
+| **Purpose** | Systematic SEO improvement |
+| **Agents** | seo-specialist, content-creator, attraction-specialist |
+| **Human Checkpoints** | Strategy approval for major changes |
+
+---
+
+## Workflow Stages
+
+### Stage 1: Audit
+
+**Agent:** `seo-specialist`
+
+**Actions:**
+1. Technical SEO audit
+2. Content audit
+3. Competitor analysis
+4. Current rankings check
+
+**Output:** SEO audit report in `plans/reports/`
+
+**Prompt Template:**
+```
+Perform SEO audit for [domain/site].
+Include:
+- Technical issues (crawl, speed, mobile)
+- Content gaps
+- Top competitor comparison
+- Current keyword rankings
+Save to: plans/reports/[date]-seo-audit.md
+```
+
+---
+
+### Stage 2: Keyword Research
+
+**Agent:** `seo-specialist`, `attraction-specialist`
+
+**Actions:**
+1. Identify topic clusters
+2. Map search intent
+3. Analyze keyword difficulty
+4. Find content gaps
+5. Prioritize opportunities
+
+**Output:** Keyword strategy doc
+
+**Prompt Template:**
+```
+Research keywords for [topic/niche].
+Include:
+- Primary keywords (high intent)
+- Long-tail opportunities
+- Question-based keywords
+- Competitor keyword gaps
+Prioritize by: volume, difficulty, intent
+```
+
+---
+
+### Stage 3: On-Page Optimization
+
+**Agent:** `seo-specialist`
+
+**Actions:**
+1. Title tag optimization
+2. Meta description updates
+3. Header structure (H1-H6)
+4. Internal linking
+5. Schema markup
+
+**Output:** Optimization recommendations
+
+**Prompt Template:**
+```
+Optimize on-page SEO for [page/URL].
+Target keyword: [keyword]
+Include:
+- Title tag suggestion
+- Meta description
+- Header optimization
+- Internal link opportunities
+- Schema markup (JSON-LD)
+```
+
+---
+
+### Stage 4: Content Creation
+
+**Agent:** `content-creator`, `seo-specialist`
+
+**Actions:**
+1. Create SEO-optimized content
+2. Build pSEO templates
+3. Update existing content
+4. Link building content
+
+**Output:** SEO content in `content/`
+
+**Prompt Template:**
+```
+Create SEO content for [keyword].
+Search intent: [informational/transactional/etc]
+Include:
+- Optimized title and headers
+- Target and related keywords
+- Internal links
+- Schema markup
+Word count: [target based on competition]
+```
+
+---
+
+### Stage 5: Monitoring
+
+**Agent:** `analytics-analyst`, `seo-specialist`
+
+**Actions:**
+1. Track keyword rankings
+2. Monitor organic traffic
+3. Analyze CTR improvements
+4. Identify new opportunities
+5. Report on progress
+
+**Output:** SEO performance reports
+
+**Prompt Template:**
+```
+Monitor SEO performance for [domain].
+Metrics:
+- Keyword ranking changes
+- Organic traffic trends
+- Top performing pages
+- CTR improvements
+Period: [date range]
+Recommend next optimizations.
+```
+
+---
+
+## SEO Checklist
+
+### Technical SEO
+- [ ] Site crawlable
+- [ ] XML sitemap submitted
+- [ ] Robots.txt configured
+- [ ] HTTPS enabled
+- [ ] Mobile-friendly
+- [ ] Page speed optimized
+- [ ] Core Web Vitals passing
+
+### On-Page SEO
+- [ ] Unique title tags (50-60 chars)
+- [ ] Meta descriptions (150-160 chars)
+- [ ] H1 tag with keyword
+- [ ] Header hierarchy (H2-H6)
+- [ ] Keyword density (1-2%)
+- [ ] Image alt text
+- [ ] Internal links
+- [ ] External links to authority
+
+### Content SEO
+- [ ] Addresses search intent
+- [ ] Comprehensive coverage
+- [ ] Original/unique value
+- [ ] Updated regularly
+- [ ] Proper length vs competition
+
+### Off-Page SEO
+- [ ] Quality backlinks
+- [ ] Brand mentions
+- [ ] Social signals
+- [ ] Local citations (if applicable)
+
+---
+
+## pSEO (Programmatic SEO)
+
+### Template Structure
+```
+/[category]/[topic]/[location]
+/tools/[tool-name]
+/alternatives/[competitor]
+/vs/[product-a]-vs-[product-b]
+```
+
+### Template Prompt
+```
+Create pSEO template for [pattern].
+Variables: [variable-list]
+Include:
+- Title template with variables
+- Meta description template
+- Content sections (dynamic + static)
+- Schema template
+- Internal linking pattern
+```
+
+---
+
+## Integration Points
+
+### Hooks Triggered
+- `campaign-tracking` → SEO content published
+
+### Data Flow
+- Audit → Research priorities
+- Research → Content strategy
+- Content → Rankings
+- Monitoring → Iteration
+
+### Tools/APIs
+- Google Search Console
+- Google Analytics 4
+- Rank tracking tools
+- Site crawlers
+
+---
+
+## Usage Example
+
+```
+/seo:audit "example.com"
+
+# Triggers workflow:
+1. seo-specialist: Technical + content audit
+2. seo-specialist: Keyword research, gap analysis
+3. seo-specialist: On-page optimization plan
+4. content-creator: New content + updates
+5. analytics-analyst: Track rankings, traffic
+```
+
+---
+
+## Success Metrics
+
+- Keyword rankings (position changes)
+- Organic traffic (sessions, users)
+- Click-through rate (impressions → clicks)
+- Conversion rate (organic traffic → goals)
+- Domain authority (backlink quality)

package/.repomixignore

@@ -0,0 +1,22 @@
+docs/*
+plans/*
+assets/*
+dist/*
+coverage/*
+build/*
+ios/*
+android/*
+tests/*
+__tests__/*
+__pycache__/*
+node_modules/*
+
+.opencode/*
+.claude/*
+.serena/*
+.pnpm-store/*
+.github/*
+.dart_tool/*
+.idea/*
+.husky/*
+.venv/*

package/AGENTS.md

@@ -1,80 +1,62 @@
 # AGENTS.md
 
-## Project
+This file provides guidance to OpenCode when working with code in this repository.
 
-PPM (Project & Process Manager) — a web-based IDE/project manager with AI chat powered by Codex Agent SDK.
+## Project Overview
 
-## Stack
+**Name:** claudekit-marketing
+**Type:** Node.js/TypeScript
+**Description:** AI-powered marketing automation toolkit built on Claude Code subagent orchestration. Enables autonomous workflows for content creation, campaign management, SEO optimization, and lead generation.
 
-- **Runtime**: Bun
-- **Backend**: Hono (HTTP) + Bun WebSocket
-- **Frontend**: React + Vite + Tailwind + shadcn/ui
-- **AI**: @anthropic-ai/Codex-agent-sdk
-- **Tests**: bun:test
+## Role & Responsibilities
 
-## Commands
+Your role is to analyze user requirements, delegate tasks to appropriate sub-agents, and ensure cohesive delivery of features that meet specifications and architectural standards.
 
-```bash
-bun dev:server    # Start backend dev (port 8081, uses ~/.ppm/ppm.dev.db)
-bun dev:web       # Start Vite frontend (port 5173)
-bun test          # Run all tests
-bun test tests/integration/  # Integration tests only
-```
-
-## Dev Config
-
-Config is stored in **SQLite** (`~/.ppm/ppm.db`). Dev uses a separate DB:
+## Workflows
 
-- **Dev**: `~/.ppm/ppm.dev.db` — port **8081**
-- **Production**: `~/.ppm/ppm.db` — port **8080**
+- Primary workflow: `./.claude/workflows/primary-workflow.md`
+- Development rules: `./.claude/workflows/development-rules.md`
+- Orchestration protocols: `./.claude/workflows/orchestration-protocol.md`
+- Documentation management: `./.claude/workflows/documentation-management.md`
+- And other workflows: `./.claude/workflows/*`
 
-`bun dev:server` automatically uses the dev database. On a new machine, run `ppm init` to create default config, then `ppm config set port 8081` for dev.
+**IMPORTANT:** Analyze the skills catalog and activate the skills that are needed for the task during the process.
+**IMPORTANT:** You must follow strictly the development rules in `./.claude/workflows/development-rules.md` file.
+**IMPORTANT:** Before you plan or proceed any implementation, always read the `./README.md` file first to get context.
+**IMPORTANT:** Sacrifice grammar for the sake of concision when writing reports.
+**IMPORTANT:** In reports, list any unresolved questions at the end, if any.
+**IMPORTANT**: Date format is configured in `.ck.json` and injected by session hooks via `$CK_PLAN_DATE_FORMAT` env var. Use this format for plan/report naming.
 
-## Release Process
+## Development Principles
 
-1. Commit feature/fix changes
-2. Update `CHANGELOG.md` with all changes
-3. Bump version in `package.json` — patch for small changes, minor/major for large ones
-4. Commit: `chore: bump version to x.x.x`
-5. Publish: `npm publish --access public`
+- **YAGNI**: You Aren't Gonna Need It - avoid over-engineering
+- **KISS**: Keep It Simple, Stupid - prefer simple solutions
+- **DRY**: Don't Repeat Yourself - eliminate code duplication
 
-## Quick SDK Tool Test
+## Documentation
 
-Use `test-tool.mjs` to verify SDK tool execution against any project cwd:
+Keep all important docs in `./docs` folder:
 
-```bash
-bun test-tool.mjs /path/to/project                    # default: echo test
-bun test-tool.mjs /path/to/project "dùng thử tool bash"  # custom prompt
+```
+./docs
+├── project-overview-pdr.md
+├── code-standards.md
+├── codebase-summary.md
+├── design-guidelines.md
+└── system-architecture.md
 ```
 
-This uses `ClaudeAgentSdkProvider` directly — same env/settings overrides as production.
-
-## Known Gotchas
-
-- **SDK .env poisoning**: Projects with `ANTHROPIC_API_KEY` in `.env` break SDK tool execution. Provider neutralizes these vars. See `docs/lessons-learned.md`.
-- **Project Codex settings**: `.Codex/settings.local.json` can restrict tools even with `bypassPermissions`. Provider overrides with empty settings.
-
-## UI Rules
-
-When creating or modifying any UI component, you MUST read and follow `docs/design-guidelines.md`, especially the **Mobile-First UI Rules** section. Key rules:
-- Dialogs → bottom sheet on mobile (below `md:` breakpoint)
-- No hover-only interactions — must have touch alternatives
-- Touch targets minimum 44×44px
-- Context menus → long-press on mobile, not tap
-- Thumb zone: primary actions in bottom 1/3 of screen for one-handed use
-- Always test both mobile and desktop layouts
+## External Files
 
-## Roadmap & Context
+Reference external instruction files in `opencode.json`:
 
-Before planning or implementing a new feature, read `docs/project-roadmap.md` to understand:
-- Which version the feature belongs to (v0.8, v0.9, v0.10, v1.0)
-- The theme and scope of that version
-- Dependencies between features
-- Strategic principles (multi-device focus, extension architecture, tiered providers)
+```json
+{
+  "instructions": ["docs/*.md", ".opencode/agents/*.md"]
+}
+```
 
-## Architecture
+---
 
-- `src/providers/Codex-agent-sdk.ts` — SDK integration, tool execution, streaming
-- `src/server/ws/chat.ts` — WebSocket chat handler
-- `src/web/hooks/use-chat.ts` — Frontend chat state management
-- `src/services/config.service.ts` — Config from SQLite (`~/.ppm/ppm.db`)
+*Generated by ClaudeKit OpenCode Generator*
+*Date: 2026-03-28*

package/CHANGELOG.md

@@ -1,5 +1,32 @@
 # Changelog
 
+## [0.9.81] - 2026-04-10
+
+### Added
+- **Persistent toast notification for pending approvals**: When an approval request or AI question arrives on a non-active session tab, a Sonner toast with "Go to session" action appears and persists until dismissed or resolved.
+
+### Fixed
+- **Approval timeout removed**: `waitForApproval` no longer auto-rejects after 5 minutes. Approvals wait indefinitely until user responds or session is cleaned up.
+- **`deleteSession` approval cleanup bug**: Was using wrong key (`sessionId` on a `requestId`-keyed map) — now properly iterates and resolves all pending approvals for the deleted session.
+- **Reconnect tool_result reliability**: Tool results are now embedded onto matching tool_use events in the buffer, so reconnecting clients don't lose tool output.
+- **Cloud WS reconnect backoff**: Reconnect attempt counter only resets after auth succeeds, preventing tight reconnect loops when server closes immediately after connect.
+- **Tunnel URL lazy sync**: `getTunnelUrl()` now falls back to reading `status.json` when supervisor sets the URL after server startup.
+
+## [0.9.80] - 2026-04-10
+
+### Fixed
+- **"Session ID already in use" error**: Removed dual-ID system (ppmId/sdkId mapping via `session_map` table). PPM UUID is now used directly as SDK session ID, eliminating the `effectiveIsFirst` heuristic that caused resume vs new-session misclassification.
+- **Non-deterministic message timing**: Replaced `setTimeout(500ms)` with `isConnected`-based pending message flush using `useRef` + `useEffect` for reliable first-message delivery after WS connect.
+- **Legacy JSON migration crash**: `config.service.ts` called removed `setSessionMapping` — replaced with `setSessionMetadata`.
+- **CLI provider session migration**: Restored `session_migrated` WS handler for CLI providers that discover session IDs from CLI output.
+- **Fork session not tracked**: Fork route now calls `resumeSession` to register forked session with provider in-memory state.
+
+### Changed
+- DB migration v16: creates `session_metadata` table, migrates data from `session_map`, cleans up orphaned ppm_id entries in `session_titles`/`session_pins`.
+- Removed `session_migrated` from `ChatEvent` type (SDK provider no longer needs it).
+- Removed `migratedSessionId` state from frontend `use-chat` hook.
+- Net deletion of ~86 lines of complexity.
+
 ## [0.9.79] - 2026-04-10
 
 ### Fixed