# Uniph.ai Feature Catalog

Last updated: February 2026

This document is the complete feature inventory for the current codebase. It is organized by product surface, API domain, automation behavior, and operator tooling.

---

## 1) User-Facing App Surfaces

### Public and Auth

- `/` - landing page with positioning, capability overview, and docs links.
- `/auth` - sign-in entry with provider-based authentication.
- `/docs` and `/docs/[slug]` - in-app documentation renderer backed by markdown files in `frontend/public/docs/`.

### Dashboard and Execution

- `/dashboard` - primary app shell for intent submission, goal management, outcomes tracking, and intent history CRUD.
- `/onboarding` - guided first-run flow for creating first outcome.
- `/outcome/[id]` - detailed outcome view.

### Goals

- `/goals/[id]` - goal detail page with status, tasks, and related execution state.

### Settings

- `/settings` - authenticated redirect to `/settings/agents`.
- `/settings/agents` - agent identity management (create, connect, refresh, revoke, audits, health).
- `/settings/channels` - channel/integration controls.
- `/settings/memory` - memory inspection and management view.

### Ops / Quality

- `/mission-control` - reliability/quality dashboard for improve metrics, release gates, harness status, and regressions.

---

## 2) Core Product Capabilities

### Intent-to-Outcome Pipeline

- User submits high-level intent text.
- Backend parses intent and creates outcomes/stages.
- Outcome progress streams to frontend via SSE.
- Outcome actions include download/share/approve/retry.

### Goals and Autonomous Routing

- Goals can be created, edited, archived, deleted, and reordered.
- Goal execution path is selected dynamically (`n8n`, `fallback`, `needs_input`) based on available capabilities.
- Routing decisions and routing results are logged for learning and future policy adaptation.
- If required input is missing, user input requests are generated and surfaced in UI.

### Intent CRUD (History Management)

- List intents with linked outcome status/progress.
- Edit intent input.
- Delete intent (optionally paired with outcome deletion at API level).

### Task and Outcome Transparency

- Stage-level status and progress updates.
- Stuck/hung detection on backend stream + frontend stale watchdog.
- Execution-path visibility cards for recently routed goals.

### Suggestion System

- Time/context/pattern/proactive suggestions.
- Dismiss and interaction tracking endpoints.
- Suggestion-driven intent creation and safe navigation actions.

---

## 3) Integrations and Channels

### Integrations

- User integrations API for provider connection metadata and encrypted config.
- Schedule context + scheduler run support.
- GitHub sync support for repo/task coordination.

### Agent Identities

- Per-agent identities with provider, scopes, status, and health checks.
- OAuth connect/refresh/revoke lifecycle.
- Audit logging for identity events.

### Channel Adapters

- Slack events endpoint.
- Discord interactions endpoint.
- Email inbound and digest endpoints.

---

## 4) API Domains (Mounted in `backend/src/app.ts`)

- `/api/users` - user sync/profile/preferences.
- `/api/goals`, `/api/tasks` - life management.
- `/api/integrations`, `/api/schedule`, `/api/github` - capability integrations.
- `/api/workspaces`, `/api/contributions`, `/api/agents`, `/api/runs` - collaboration fabric and execution contexts.
- `/api/contracts`, `/api/governance`, `/api/execution`, `/api/evaluation`, `/api/memory`, `/api/orchestrate` - governed execution + quality systems.
- `/api/runtime/*` - modern intent/outcome/goal/suggestion/workflow routes for dashboard runtime.

V2 highlights:

- `/api/runtime/intent` (submit intent)
- `/api/runtime/intents` (intent CRUD)
- `/api/runtime/outcomes` and `/api/runtime/outcomes/:id/progress` (outcome list/detail/SSE)
- `/api/runtime/goals` (goal lifecycle + execution path kickoff)
- `/api/runtime/user-input-requests` (clarification workflow)
- `/api/runtime/workflows`, `/api/runtime/webhooks` (n8n templates/executions/callbacks)

---

## 5) Observability, Learning, and Quality Gates

- Evaluation checkpoints for route selection, outcome lifecycle, suggestion interactions, and improvement pulses.
- Improvement pulse loop analyzes regressions and tracks runtime status.
- Mission Control expose:
  - friction metrics
  - release gate checks
  - onboarding harness
  - routing learning sample/failure/confidence/policy thresholds.

---

## 6) Security and Access Controls

- Session-based user auth middleware and protected route checks.
- Agent API-key authorization for agent-specific routes.
- Protected frontend route gate via proxy middleware for dashboard/goals/settings.

---

## 7) What Was Refactored in This Pass

- Removed repeated goal-normalization logic in `V2DashboardClient` and centralized it into helpers.
- Removed unused `TaskProgressIndicator` import from `V2DashboardClient`.
- Removed `any` usage in runtime goal create flow by passing a typed create payload.
- Kept behavior unchanged while reducing duplicate logic and maintenance risk.