ARCANADA
← Ecosystem / ARCA Assistant
PLANNING L1 · target L3

ARCA Assistant

Single point of entry into the entire ecosystem.

Every Arcanada service has its own entry point. ARCA Assistant is planned as a conversational front-end that would route user intent to the right one: ask Verdicus to translate, Transcribator to convert audio, Munera to track a task.

It is in the planning phase, with target port 3800. It is distinct from Argana AI (the commercial product): ARCA Assistant is planned as the operator-facing ecosystem control plane, sharing the @arcanada/core libraries.

Status
Planning
Autonomy (AAL)
L1 → L3
Target port
3800
Shared code
@arcanada/core

01

Say it once

Natural-language requests are meant to be routed to the right Arcanada service, or to several in one go.

02

Every action has an owner

Each action is meant to run in an Auth Arcana session, with an audit trail of every routed intent.

03

Not Argana

ARCA Assistant serves the operator; Argana AI is the separate commercial product, planned on the same @arcanada/core.

Planned capabilities

Planned

  • Natural-language intent routing across all Arcanada services
  • Cross-service orchestration (one prompt → multiple service calls)
  • Auth Arcana session — every action is identity-attached
  • Per-user memory via LTM (recall preferences across sessions)
  • Web and Telegram clients
  • Shared @arcanada/core with Argana AI (DRY connector layer)
  • Audit trail of every routed intent for post-hoc review

Current autonomy level

L1
What levels mean →

Weakest link

Not yet in production. The L4 self-healing capabilities of downstream services do not help if ARCA Assistant itself misroutes intents — intent-classification evaluation set is the gating artifact.

Roadmap to L3

  1. L1 → L2 — MVP launch with manually curated intent map; deploy reporting to Ops Bot.
  2. L2 → L3 — observability stack, intent-classification eval set (≥500 labeled samples), fallback to fail-fast on unknown intent rather than guessing.
  3. Verification gate — eval set ≥ 95% accuracy and zero silent misroutes for 30 days post-launch.

Next step

Review the plan and future launch gates

PLANNING does not mean availability or active development. The roadmap above is the next reference; no launch date is claimed.

Review the path to the target level