Skip to main content

Positioning Design: The Hosted OpenClaw Alternative for Teams and Products

Date: 2026-04-25 Status: Draft — pending user approval Owner: VG Scope: Positioning reset for Distri, cascading into the GitHub README, the docs site (distri.dev/docs), and a separate web spec for distri.dev landing.

TL;DR​

Distri repositions as "The hosted OpenClaw alternative for teams and products."

  • Wave-riding anchor: OpenClaw is the viral (247k-star, Mar 2026) open-source personal AI agent. We slot Distri into the same category but for a different audience and deployment model.
  • Wedge: OpenClaw is personal, local, one-laptop. Distri is team, hosted, embeddable — same markdown-first, open-core ethos, different surface.
  • Three modes (primary IA): Channels → Embed → Local/OSS.
  • Rollout: README first, docs second, website spec third (implemented in a separate session).

Context​

What OpenClaw is (and why it matters now)​

OpenClaw is an open-source autonomous AI agent (MIT) that uses messaging platforms as its primary UI — WhatsApp, Telegram, Slack, Discord, Signal, iMessage, Matrix, ~20 more. Agents and skills are stored as plain Markdown + YAML in a workspace folder (~/.openclaw, SOUL.md). It has a skills marketplace ("ClawHub") and an enterprise wrapper from NVIDIA (NemoClaw).

Key facts driving the positioning:

  • ~247k GitHub stars by March 2026. Genuinely viral.
  • Local-first and self-hosted — lives on the user's laptop.
  • Chat-app UX is the hero feature.
  • Foundation-governed (Peter Steinberger moved to OpenAI; nonprofit foundation established Feb 2026).
  • Security surface is broad (email, calendars, shell) — enterprise buyers are wary.

What Distri already has (verified in code)​

  • Channels — distri-cloud/distri-gateway/ ships webhook adapters for Telegram, WhatsApp, Discord, Slack. distri-types/channels.rs models Bot, Channel, BotConnection, AuthenticatedChannelUser, PlatformAuthScope (workspace-scoped vs open), TriggerMode, channel verification / OAuth gating.
  • Embed — distri-cloud/distri/distrijs — React SDK with <Chat />, SSE streaming, tool renderers, voice, developer mode, IndexedDB-backed browser tools.
  • Hosted cloud — app.distri.dev, full dashboard (screenshots in distri.dev/static/img/page_*.png).
  • Open core / local — MIT root + Elastic License 2.0 on server/; brew install, curl | sh, SQLite, no external deps.
  • Agents-as-markdown + skills + sub-agents + MCP + A2A — feature parity with OpenClaw on the core agent-runtime axes, with deeper support for team-scoped deployment.

All three modes are shipping today. Positioning can be declarative, not aspirational.

The Positioning​

The headline (locked)​

The hosted OpenClaw alternative for teams and products.

Consistent across README hero, docs intro, web H1, and meta description.

The sub (one canonical long form, abbreviated per surface)​

Workspace-scoped agents on Slack, WhatsApp, Discord, and Telegram — or embedded directly in your product. Gated by your own OAuth connections. Open core, run it locally if you want.

The wedge (one sentence)​

OpenClaw runs on your laptop for you. Distri runs in the cloud for your team — and inside your users' products.

Audience wedge​

DimensionOpenClawDistri
Who it's forIndividuals, hobbyists, prosumersTeams, companies, product builders
Where it runsThe user's laptop (self-hosted)Hosted (default) or self-hosted
Primary UXPersonal chat appsWorkspace-scoped channels + in-product embed
Auth modelThe user's own accountsOAuth connections gated per workspace
ExtensibilitySkills marketplace (ClawHub)Skills + tools + MCP + React SDK
FoundationOpenClaw Foundation (nonprofit)Distri (VC-backed, open-core commercial)

The three modes (primary IA)​

  1. Channels — Workspace-scoped bots on Slack / WhatsApp / Discord / Telegram. Slack and Discord inherit platform workspace auth; Telegram/WhatsApp are gated by distri-managed OAuth connections. Trigger modes (all, on_mention), per-channel verification, /switch workspace selection on system bots.
  2. Embed — @distri/react → drop <Chat /> into your product. SSE streaming, tool renderers, voice, developer mode. Browser-side tools (Read/Write/Edit/Glob/Grep) backed by IndexedDB.
  3. Local / Open Core — brew install distri, distri serve, SQLite, markdown agent definitions. MIT root, Elastic License 2.0 on server — no vendor lock-in.

These three are the top-level IA in README, docs, and website.

Wave-riding tactics​

We name OpenClaw explicitly and repeatedly — not as a competitor, as a category anchor. Supabase built its early positioning on "the open-source Firebase alternative." Same move.

Where we name OpenClaw:

  • H1 — The hosted OpenClaw alternative for teams and products.
  • Meta description (SEO) — same sentence, verbatim. Captures "OpenClaw alternative" search queries.
  • README first paragraph — one sentence comparing the two.
  • Docs intro — one sentence comparing the two.
  • Web page — a compact comparison card (OpenClaw vs Distri) below the fold.
  • Launch tweet / social — "OpenClaw, for the rest of your company." as variant copy.

Where we do not name OpenClaw:

  • CLI output, error messages, product UI.
  • Product names — Distri is Distri. No "Claw" riff in naming.
  • Legal copy.

Trademark posture​

"OpenClaw" is used comparatively and nominatively — the standard US-law fair use for identifying another product in marketing. This is the exact move that "the open-source Firebase alternative" (Supabase) made. No use of OpenClaw logos, no product name collisions, no implied affiliation.

Milestone 1 — Repository README​

File: ../distri-cloud/distri/README.md

Current problem: Leads with "open-source AI agent framework built in Rust" and "Define agents in markdown, connect your product functions as tools". Technically accurate but buries the story:

  • Doesn't name the category (OpenClaw).
  • Doesn't name the three delivery modes.
  • Channels are entirely absent — a major capability.
  • Reads as a framework, not a product.

Target structure​

  1. Hero — new headline + sub + three-mode pitch.
  2. Three modes — one block each: Channels, Embed, Local. Each shows the minimum "hello world" code path.
  3. Quickstart — brew install → push an agent → choose a delivery mode.
  4. Agent definition format — keep, shorten.
  5. Built-in agents — keep table.
  6. Skills — keep, trimmed.
  7. React SDK — keep but reframe as "Embed mode".
  8. Architecture — move below the fold, keep for engineers.
  9. Licensing / links — keep.

Images to refresh​

Existing images in distri.dev/static/img/ are cloud dashboard screenshots. For the new README:

  • Hero image — needs a new composite showing: a Slack/WhatsApp conversation with a bot on the left, a React <Chat /> embed in the middle, a terminal on the right. (TODO: image production — flagged as a follow-up; first pass of README can use page_home.png and a separate channels screenshot if one exists in distri-cloud/.)
  • Channels section — screenshot of the "Bots" dashboard page.
  • Embed section — existing embedded.png or page_agent_embed.png.
  • Local section — terminal screenshot.

Milestone 2 — Docs restructure​

Root: /Users/vivek/projects/distri/distri.dev/docs/

Current IA:

intro.mdx
getting-started.mdx
samples.mdx
theming.mdx
concepts/
guides/
cloud/
self-hosted/
reference/
tutorials/
managing-agents.mdx
lifecycle-events.mdx

Problem: Organized by engineering surface (concepts, guides, cloud, self-hosted), not by the buyer's mode of use. A reader who wants to "build a Slack bot" has no landing path.

Target IA​

intro.mdx                    # New positioning intro, three-mode pitch
getting-started.mdx          # Hosted-first quickstart
modes/
  channels/                  # NEW — first-class section
    index.mdx                # Overview, which platforms, when to use
    slack.mdx
    discord.mdx
    telegram.mdx
    whatsapp.mdx
    oauth-gating.mdx
    trigger-modes.mdx
  embed/                     # Renamed/promoted from guides/client
    index.mdx
    react-sdk.mdx
    streaming.mdx
    browser-tools.mdx
    voice.mdx
  local/                     # Renamed from self-hosted
    index.mdx
    install.mdx
    server.mdx
    sqlite.mdx
concepts/                    # Keep — agent definition, tools, skills, threads
agents/                      # Merged managing-agents + built-in agents
skills/                      # Promoted to top-level
tools/
  builtin.mdx
  dynamic-http.mdx
  external.mdx
  mcp.mdx
samples/                     # Keep
reference/                   # Keep — CLI, API, config

Content changes​

  • intro.mdx — replace hero paragraph with the new headline + wedge; add OpenClaw comparison block; reshape "Three Steps to Ship" into "Three Modes" (channels / embed / local).
  • getting-started.mdx — change from generic "push an agent" to "pick your mode" branching.
  • Migrate guides/client/* → modes/embed/*.
  • Migrate self-hosted/* → modes/local/*.
  • Net new: modes/channels/* (currently no docs for a shipping feature).

Milestone 3 — Website positioning spec​

Scope note: This milestone produces a spec document only. The spec is written in this session (after README and docs are done). Website code/content changes happen in a separate session with its own writing-plans pass, using this spec as the single source of truth.

Output file (this session): docs/superpowers/specs/2026-04-25-website-positioning-spec.md — written after Milestones 1 and 2 land.

Website spec will cover​

  • Hero — H1, sub, primary CTA ("Start free — no server needed"), secondary CTA ("Read the docs").
  • Wave-rider strip — One-line comparison block immediately below the hero: "OpenClaw lives on your laptop. Distri lives at work." with a tiny comparison table (3–4 rows).
  • Three-mode cards — Channels, Embed, Local. Each with a 30-word pitch, one code snippet or screenshot, and a CTA into the matching docs mode.
  • Proof section — Who's building with Distri (logos / testimonials if available), or a production-grade strip ("Rust runtime · SSE streaming · SQLite · A2A · MCP").
  • Quickstart accordion — The three modes' minimum hello-world paths inlined.
  • Why Distri vs OpenClaw — a dedicated section with the wedge table.
  • Open core trust strip — license breakdown, GitHub link, "run it locally" CTA.
  • Footer — docs, cloud, GitHub, socials, pricing.

What the website spec will NOT cover (out of scope for this project)​

  • Full site IA (pricing page, blog, changelog) — keep current.
  • Design-system revamp — reuse current components.
  • New illustrations / marketing renders — flag as follow-up.

Non-goals​

  • Renaming Distri or any product (e.g. to "Claw"-something). We ride the wave via copy, not naming.
  • Sunsetting the Rust/framework angle. It stays — as proof below the fold, not the pitch above it.
  • Deprecating self-hosted / local mode. It stays a first-class mode (the "open core" column of the three-mode IA).
  • Rebuilding the distri.dev site in this project. Spec only; implementation is a separate session.

Risks and open questions​

  1. OpenClaw wave longevity. If OpenClaw cools or the foundation stumbles, the H1 ages badly. Mitigation: audience-side wedge ("for teams and products") still reads right even if "OpenClaw" gets swapped out for the next category leader; the sub and three-mode IA stand alone.
  2. Trademark posture. Comparative/nominative use is standard, but OpenClaw foundation policy is unknown. Mitigation: no logos, no implied affiliation; only plain-text nominative reference in marketing copy. If a cease-and-desist lands, fall back to "the hosted open-core agent platform for teams and products."
  3. Channel infra maturity. Code ships for Telegram/WhatsApp/Discord/Slack, but operational maturity (rate limits, delivery guarantees, error paths, provider-side onboarding) varies per platform. Mitigation: docs should be candid per-platform and not overclaim; any platform not yet GA should carry a beta tag.
  4. Image assets. The hero composite image (chat app + React embed + terminal) doesn't exist yet. Mitigation: Milestone 1 ships with a best-available subset; flag image production as a follow-up task.
  5. "Alternative" framing risks looking derivative. Some buyers will read "alternative" as "lesser knockoff." Mitigation: the sub and wedge explicitly reframe it as a different-audience play, not a cheaper substitute.

Success criteria​

  • A reader hitting the README, the docs intro, or the website homepage can answer in 10 seconds:
    1. Who is this for? (teams building products)
    2. What is this like? (OpenClaw)
    3. How is it different? (hosted, workspace-scoped, embeddable)
    4. Where do I start? (three modes)
  • "OpenClaw alternative" search query lands on distri.dev within 30 days of web launch.
  • Consistent language across README, docs, and web — no contradictions.

Rollout plan (for the implementation phase)​

The writing-plans skill will turn this into a detailed task plan. High-level sequence, all in this session:

  1. Milestone 1: README — single-file edit at ../distri-cloud/distri/README.md. Quick feedback loop.
  2. Milestone 2: Docs — IA move + content updates under distri.dev/docs/. Multi-file.
  3. Milestone 3: Website spec — produce 2026-04-25-website-positioning-spec.md. Spec only. Actual website implementation is a separate session that consumes the spec.

Each milestone ends with a commit and a user review gate before moving to the next.