Skip to content

starter-pack-chooser.md

Latest commit

 

History

History
126 lines (99 loc) · 5.93 KB

starter-pack-chooser.md

File metadata and controls

126 lines (99 loc) · 5.93 KB
Raw

WebaiBridge Starter Pack Chooser

If the only sentence in your head is:

Which starter pack should I pick first?

this page is the answer.

In plain English:

  • starter-packs/README.md is the shelf overview
  • starter-packs/index.json is the warehouse map
  • this page is the front-desk guide

It does not invent new capability. It reorganizes the packs that already exist into a cleaner first decision.

Machine-Readable Source

Read-Only Access

pnpm run webai-bridge:cli -- starter-pack-chooser
pnpm run webai-bridge:cli -- starter-pack-chooser-schema
pnpm run webai-bridge:cli -- starter-pack-scenario --target codex-builder
pnpm run webai-bridge:cli -- starter-pack-scenario --target docs-seo-sync-skill
pnpm run webai-bridge:cli -- starter-pack-scenario --target chat-app-runtime-skill
  • webai-bridge.catalog.starter_pack_chooser
  • webai-bridge.catalog.starter_pack_chooser_schema
  • webai-bridge.catalog.starter_pack_scenario

Quick Pick

If your first job is... Choose this pack Why Do not expect
Bridge Codex-style text requests into WebaiBridge codex thin Responses-style runtime bridge tool execution parity / MCP parity / worktree parity
Bridge Claude Code-style message payloads claude-code thin message/runtime bridge terminal shell parity / approval parity / tool parity
Keep OpenClaw-style delegation without copying the product shell openclaw delegation-first bridge with product boundary intact operator parity / product-shell parity
Inspect runtime truth through MCP mcp read-only runtime inspector over stdio execution brain / write plane
Build a provider triage recipe runtime-diagnostics-pack read-only diagnostics and support-bundle recipe invoke / acquisition write / browser automation
Sync docs or SEO wording to truthful labels docs-seo-sync-pack discoverability helper with human review built in marketing autopilot / launch automation
Start a chat app runtime bridge without promising a full chat shell chat-app-runtime-pack service-first chat runtime starter with a bounded handoff route full chat product scaffold / tool-using shell
Ground a research copilot in truth surfaces before synthesis research-copilot-pack truth-grounded research starter with human review still in the loop autonomous research loop / citation autopilot
Compare runtime routes before choosing a winner compare-runtime-pack compare-first worksheet that keeps tradeoffs reviewable automatic winner selection / benchmark autopilot
Stay BYOK-first with safe invoke boundaries before touching Web/Login byok-first-safe-pack safe lane-first starter for provider truth and invoke planning key storage workflow / web-login recovery / failover promises

Decision Flow

flowchart TD
    A["What are you building first?"] --> B["A builder/helper"]
    A --> C["A skill or automation pack"]
    B --> D["Codex-style text bridge"]
    B --> E["Claude Code-style message bridge"]
    B --> F["OpenClaw-style delegation bridge"]
    B --> G["Read-only MCP inspector"]
    C --> H["Provider triage recipe"]
    C --> I["Docs or SEO truth-sync recipe"]
    C --> P["Chat app runtime bridge starter"]
    C --> Q["Research copilot grounded in truth surfaces"]
    C --> R["Compare runtime routes before choosing"]
    C --> S["BYOK-first safe starter"]
    D --> J["Choose `codex`"]
    E --> K["Choose `claude-code`"]
    F --> L["Choose `openclaw`"]
    G --> M["Choose `mcp`"]
    H --> N["Choose `runtime-diagnostics-pack`"]
    I --> O["Choose `docs-seo-sync-pack`"]
    P --> T["Choose `chat-app-runtime-pack`"]
    Q --> U["Choose `research-copilot-pack`"]
    R --> V["Choose `compare-runtime-pack`"]
    S --> W["Choose `byok-first-safe-pack`"]
Loading

What This Page Helps With

  • first-time builder onboarding
  • plugin/skills/automation discoverability
  • SEO pages that answer a real route-selection question
  • machine-readable pack selection for local tooling

Use-Case Skill Packs

These four are now front-door use-case skill packs.

They are still bounded, still partial, and still not full product shells. But they are no longer hidden as "maybe later" shelf notes. If one of these is your first job, it is a legitimate first-row route:

Use case Pick this pack First handoff after choosing
Service-first chat runtime bridge chat-app-runtime-pack pnpm run webai-bridge:cli -- skill-pack-route --target chat-app-runtime-pack
Truth-grounded research copilot research-copilot-pack pnpm run webai-bridge:cli -- skill-pack-route --target research-copilot-pack
Compare-first runtime or pack selection compare-runtime-pack pnpm run webai-bridge:cli -- skill-pack-route --target compare-runtime-pack
BYOK-first safe invoke planning byok-first-safe-pack pnpm run webai-bridge:cli -- skill-pack-route --target byok-first-safe-pack

Once you pick one of these packs, move to docs/host-integration-playbooks.md for the coordinated CLI + MCP handoff, then open examples/hosts/README.md only if you need a runnable host-local example.

What It Does Not Mean

This chooser is not:

  • a plugin marketplace
  • full Codex parity
  • full Claude Code parity
  • full OpenClaw parity
  • an MCP execution brain
  • launch automation

Related Pages