Skip to main content

OpenClaw vs Seraph — Feature Gap Analysis

Date: 2026-02-09 (updated) OpenClaw version context: v2026.1.x (145k+ GitHub stars, formerly Clawdbot/Moltbot) Seraph branch: develop (Phase 1 + 2 complete)

Overview

OpenClaw is a self-hosted gateway connecting messaging platforms (WhatsApp, Telegram, Discord, etc.) to AI agents. It's headless — text-in/text-out with no visual UI.

Seraph is a self-contained web app with a retro 16-bit RPG village UI. A Phaser 3 canvas renders a tile-based village where an animated pixel-art avatar walks between tool stations while the user chats via an RPG-style dialog box. Persistent identity, long-term memory, hierarchical goals.

Different philosophies, but many of OpenClaw's features are worth adopting.


What Seraph Has (Phase 0-2 Complete)

  • Real-time chat with AI agent (WebSocket streaming with step/final/error/proactive/ambient types)
  • Tool execution with visual feedback (animated RPG avatar walks to tool stations in village)
  • 16 auto-discovered tools + MCP integrations: web search, file I/O, template fill, soul view/update, goal CRUD, shell execute (snekbox sandbox), browser automation (Playwright), calendar (Google), email (Gmail), task management (Things3 via MCP)
  • Persistent sessions — SQLite-backed, survive restarts, session list UI with switch/delete
  • Persistent memory — Soul file (soul.md) + LanceDB vector store with sentence-transformer embeddings
  • Memory consolidation — Background extraction of facts/preferences/decisions after each conversation
  • Hierarchical goal system — Vision → Annual → Quarterly → Monthly → Weekly → Daily, with quest log UI
  • Onboarding flow — Specialized agent for first-time users, skip/restart controls, welcome message
  • Plugin system — Auto-discovery of tools from src/tools/, tool registry with village metadata
  • Sandboxed execution — snekbox Docker sidecar for shell commands, Playwright for browser
  • Phaser 3 village scene with 7 buildings, day/night cycle, idle wandering, 12 waypoints, speech bubbles
  • Multi-model support via OpenRouter/LiteLLM
  • Docker Compose dev environment (3 services: backend, frontend, sandbox)
  • React 19 + Vite 6 + TypeScript + Tailwind + Zustand + Phaser 3 frontend

Feature Gap Analysis

Tier 1 — Remaining Critical Gaps

#FeatureOpenClawSeraph Status
1Model fallbacksPrimary + fallback chain, per-agent model override, provider rotationSingle model via OpenRouter, no fallback
2Tool policy systemAllow/deny lists per agent, profiles (minimal/coding/messaging/full), elevated modeOnboarding agent has restricted tools, but no general policy system
3Context managementContext pruning (off/adaptive/aggressive), session compaction/summarizationUnbounded history, no compaction

Tier 2 — Major Gaps

#FeatureOpenClawSeraph Status
4Proactive heartbeatCron-like scheduled tasks (check email, RSS, summaries)WS protocol scaffolding exists (proactive/ambient types), but no scheduler or reasoning engine yet (Phase 3)
5Multi-channel messagingWhatsApp, Telegram, Discord, Slack, Signal, iMessage, Mattermost, Google ChatWeb UI only
6Multi-agent routingMultiple isolated agents per gateway, deterministic routingSingle agent (+ onboarding agent)
7SubagentsSpawnable child agents with concurrency limits, agent-level restrictionsNone
8Note-taking / Knowledge baseN/A (not an OpenClaw feature)Planned in roadmap (Phase 2.4) but not implemented — no Obsidian/markdown vault integration

Tier 3 — Important Gaps (UX & operational)

#FeatureOpenClawSeraph Status
9Streaming/chunkingBlock streaming with configurable chunk size, human-like delayRaw WebSocket step streaming
10Media supportSend/receive images, audio, documents bidirectionallyText only
11TTSElevenLabs/OpenAI providers, auto/inbound/tagged modesNone
12Voice transcriptionInbound voice note transcription hookNone
13Message queuingSteer/followup/collect/interrupt modes, debouncing for rapid messagesNo queue, one-at-a-time
14Security audit CLIopenclaw security audit --deep, permission hardening, log redactionNone
15User auth/identityDM pairing, allowlists, identity links across channels, access groupsAnonymous singleton user, no auth
16Configuration UIWeb control dashboard, config editing via chatNo settings UI
17Remote accessSSH, Tailscale, mDNS discoveryLocalhost only
18Structured loggingRedaction, pretty/compact/json styles, per-file outputBasic console logging

Tier 4 — Nice-to-Have

#FeatureOpenClaw
19Mobile nodes (iOS/Android with Canvas)
20macOS menubar companion app
21Group chat mention gating & policies
22Config includes with deep merge (10 levels)
23Response prefix templates ({model}, {identity.name})
24Ack reactions (emoji confirmations)
25Custom chat commands (/command in chat)

Previously Identified Gaps — Now Resolved

These were gaps in the original analysis that have since been implemented:

FeatureOriginal GapResolution
Persistent memoryIn-memory only, lost on restartSoul file + LanceDB vector store (Phase 1)
Session persistenceIn-memory dict, no persistenceSQLite-backed sessions with full history (Phase 1)
Sandboxed executionNo sandboxingsnekbox Docker sidecar (Phase 2)
Browser automationDuckDuckGo text search onlyPlaywright with headless Chromium (Phase 2)
Shell command executionNo shell toolsnekbox-based sandboxed execution (Phase 2)
Plugin/skill system4 hardcoded toolsAuto-discovery from src/tools/ (16 native tools) + MCP integrations (Phase 2)

Phase 1 — Foundation (make the agent robust) — DONE

  1. Persistent sessions + chat history — SQLite, survive restarts
  2. Persistent memory system — Agent "soul" / long-term recall across sessions
  3. Context management — Compaction/summarization for long conversations
  4. Model fallback chain — Primary + fallback models, graceful degradation

Phase 2 — Capability Expansion — DONE

  1. Shell execution tool — With sandboxing/allowlists
  2. Browser automation tool — Playwright-based, huge capability unlock
  3. Media support — Image send/receive in chat
  4. Plugin/skill system — User-installable tools without backend code changes

Phase 3 — Operational Maturity

  1. Tool policies — Allow/deny per session or user
  2. Security sandboxing — Docker-based tool execution (partially done via snekbox)
  3. Settings UI — In-app configuration panel
  4. User auth — Basic identity + session isolation

Phase 4 — Distribution & Polish

  1. Telegram/Discord bot — Alternative frontends leveraging existing backend
  2. Scheduled/proactive tasks — Heartbeat system for autonomous workflows
  3. TTS — Fits the RPG theme (avatar "speaking" with voice)
  4. Structured logging — Redaction, multiple output formats

Seraph's Unique Advantage

OpenClaw is headless. Seraph's visual RPG experience has no equivalent:

  • Phaser 3 village scene with 7 buildings mapped to tool categories
  • Animated pixel-art avatar walking between tool stations
  • Day/night cycle based on system time
  • Idle wandering between 12 waypoints
  • Speech bubbles with step content
  • Quest log UI with hierarchical goals and domain progress
  • Persistent identity and onboarding that builds a relationship
  • CRT scanline/vignette retro effects

Sources