View memory

Instructions for AI Assistants

Current state: Memory loaded
Next action: Read the memory content above. To search more, fetch /recall.
Fetch this URL: /memory/283cd5d0-79dc-48e4-a86a-106915c2212d

Memory

Looked at operator-shared estate opencode-stack/slim-chat (read-only scratch clone, nothing touched, no services contacted). Shape: monorepo — apps/chat (Next.js chat UI) + ops-console, packages/tms (opencode plugin), workers/voice (CF Agents voice), server (Total.js, native TMS :18789), hub :8080 (7 apps), memory CMS :8003, Flow :8081, OpenFiles :8082, OpenLogger :8083, ActivePieces :8085 ("caged", OAuth loop closed, 9th hub app), OpenBao :8200 dev (kernel/ KV incl. redaction token vault tokens/). Mapping to tonight's stack: this estate = EXECUTION plane (already real: conductor, vault home, chat surface, memory, TMS); new estate = GOVERNANCE/product plane (rules, provenance, funnel contracts, wrapper curation). Validations: vault has a concrete home (OpenBao kernel/tokens); "caged" AP matches wrapper curation; native TMS server matches divert transport. Open strategic point (not decided): own-chat-UI vs meet-in-their-LLM — read as complementary (owned = embedded/workspace surface, external = distribution funnel), operator to confirm.

[Digest: 53df278b-50a4-44ed-a3d2-156dfa06cc37]

Tags: opencode-stack, execution-plane, mapping, project:redaction-kernel, kind:episodic, rolled-up — Source: mcp — 2026-10-08 15:13:31 UTC

What would you like to do next?