PostQode User Onboarding Guide

Welcome to PostQode! This onboarding guide takes you step-by-step from account sign-up and email activation to choosing your account type (Personal vs. Organization), installing the extension, and running your first AI-assisted task.

---

1. Account Sign-Up & Email Activation

Creating a PostQode Account

1. Go to PostQode Sign Up (or click Sign Up directly within your IDE Extension). For a detailed walkthrough, see Setting up PostQode.
2. Enter your Full Name, Work Email, and Password.
3. Click Create Account.

Email Activation & Verification

After submitting the sign-up form:

1. PostQode sends an activation email to your registered email address.
2. Open the email and click Activate Account (or enter the verification code).
3. Once verified, you will be redirected to the PostQode Web Console.

[!IMPORTANT] Joining an Organization via Invitation: If you were invited by your organization or team administrator, click the Accept Invitation link in your email invite to complete registration and automatically join your designated Organization and Team.

---

2. Personal vs. Organization Accounts

PostQode supports two distinct account contexts:

                            User Identity
                                 │
         ┌───────────────────────┴───────────────────────┐
         ▼                                               ▼
  Personal Account                             Organization Account
  • Assigned automatically on sign-up          • Managed enterprise tenant
  • For individual projects & BYOK keys         • Shared Gateway Proxy & billing
  • Isolated custom rules & skills             • Role-Based Access Control (RBAC)
                                               • Team budgets & model governance

Account Comparison

Feature	Personal Account	Organization Account
Primary Purpose	Personal coding, solo projects, experimentation.	Team collaboration, enterprise projects, production deployments.
API Key Management	Direct personal API keys (OpenAI, Anthropic, OpenRouter, local Ollama).	Centralized Gateway Proxy keys managed by Org Admins.
Budgeting & Limits	User-managed billing or free models.	Enterprise bulk budgets, soft/hard spending caps set by Org Admins.
Governance & Rules	Personal custom rules & skills (~/.postqode/).	Organization-wide model policies, team rules (.postqode/), and SCIM sync.

How to Switch Between Accounts

• In the PostQode Dashboard/App (app.postqode.ai): Click the account selector in the top-left sidebar to switch between Personal Workspace and your Organization Workspace(s).
• In the VS Code Extension: Open the Accounts icon in the sidebar / title bar ➔ Select the Organization or Personal account from the dropdown.

---

3. Installation & First Launch

Installing the IDE Extension

1. Open VS Code (or your preferred IDE).
2. Go to the Extensions tab (Ctrl+Shift+X / Cmd+Shift+X).
3. Search for PostQode.
4. Click Install.
5. Once installed, click the PostQode icon () in the Activity Bar to open the PostQode panel.

---

4. Choosing Your Provider & Authentication

When launching PostQode for the first time, you can choose how PostQode connects to AI models:

• Option A: PostQode Provider (Recommended for Managed Access): Select PostQode Provider in settings to use managed AI access routed through the PostQode Gateway Proxy. No individual model API keys required.
• Option B: Bring Your Own API Key (BYOK): Select individual AI providers (e.g., Anthropic Claude, OpenAI, OpenRouter, DeepSeek, Google Gemini, or local models via Ollama/LM Studio) and enter your secret API key.

---

5. Initializing Your Workspace (AGENTS.md)

Before running your first task, initialize workspace guidelines so PostQode understands your project structure, tech stack, and coding standards:

1. Open the PostQode chat panel.
2. Type /init and press Enter.
3. PostQode will scan your codebase, analyze your dependencies, ask clarifying questions, and generate an AGENTS.md file in your project root.

---

6. Understanding Modes & Built-in Agents

PostQode provides two primary execution modes alongside specialized built-in agents tailored to different phases of the software development lifecycle:

Primary Execution Modes

Mode	Primary Purpose	Best Used For
Plan Mode	Read-only exploration, architecture design, and interactive alignment	Analyzing codebases, drafting implementation plans, discussing design tradeoffs
Agent Mode	Execution, code writing, and terminal command execution	Implementing features, refactoring, fixing bugs, running test commands

Built-in Specialized Agents

PostQode includes 11 packaged agents with dedicated system prompts, skills, and tools:

Built-in Agent	Description	Scoped Capabilities & Skills
code-reviewer	Reviews code diffs for bugs, security vulnerabilities, and performance	Scans OWASP vulnerabilities, traces execution, rates severity
deploy-agent	Release management and automated deployment	Runs linters, bumps versions, merges PRs, executes canary checks
design-agent	Visual UI design, styling consistency, and mockup implementation	Audits spacing/alignment, generates responsive CSS/HTML, captures UI diffs
doc-writer	Architecture, README, API docs, and changelog synchronization	Scans stale docs, writes structured docs, generates user-facing release notes
explore	Fast, read-only codebase exploration and dependency mapping	Zero-risk code search, architecture navigation, and file indexing
perf-monitor	Benchmarks page load times, Core Web Vitals, and network waterfalls	Compares performance against baselines, detects N+1 bottlenecks
product-strategist	Product strategy, feature scoping, and user journey definition	Scopes MVPs, creates product briefs, analyzes feature tradeoffs
security-auditor	Dependency supply chain audit, credentials leak scan, and threat modeling	STRIDE threat modeling, CVE scans, hardcoded secrets archaeology
web-automator	Automated browser flows, form interactions, and element locator verification	Drives browsers, generates Playwright scripts, validates UI state
web-qa	End-to-end quality assurance testing and bug reproduction	Verifies form submissions, catches JavaScript console errors, files bug reports

---

7. Running Your First Task

1. Start a Task: Type /deep-planning Create a user profile update API endpoint with Zod validation.
2. Use Context Mentions (@): Mention files (@src/routes/user.ts), git diffs (@git), or linter errors (@problems).
3. Review & Approve: Inspect proposed file edits and approve actions.
4. Explain & Commit: Type /explain-changes to review the git diff before committing code!