d2f749d33d9ba335897e90b8c24b88da11d81ce8
The root cause behind "the agent stops at the first error and doesn't recover":
provisioning executions run ASYNCHRONOUSLY (pct_create fires the SSH work in a
goroutine and returns "running" immediately), so the agent's turn ENDS before
the result exists. The agent literally isn't running when the step fails — it
can't react to a failure it never observes. The only thing that fed results
back was the operator typing "continue" after every async step: the human was
the event loop. (In the flagged 18-message session the operator typed
continue/proceed/?? eight times while the agent correctly diagnosed each failure
but couldn't advance a step on its own.)
This makes the system the event loop instead:
- migrations/017: nomos_plan_executions links each gated execution to the chat
session that started it.
- cmd/nomos: after a tool result, any "execution <uuid>" it started is linked
to the session. A background worker (continue.go) polls for those executions
reaching a terminal state and — while the agent has an open assent window (an
approved plan is in flight) — re-invokes the agent with the result
("execution X completed/failed: <result>"), so it proceeds to the next step
or diagnoses+fixes the failure, with no operator tick. Guarded against loops
(mark-continued before running) and bounded by the 30-min window.
- chatWith(): chat() variant that injects the finished-execution note after
replayed history without persisting a fake user turn.
- DecideApproval: approving a step by ANY route (button or chat-assent) now
opens the assent window, so auto-continuation works regardless of how the
operator approved — previously only typing "go ahead" opened it.
- SOUL: the agent is told it will be auto-re-invoked when async steps finish —
don't poll get_execution_status, don't wait for "continue"; end the turn and
keep going step by step until the goal is verified or a genuine blocker.
This is the root fix, not another per-command patch: you can't enumerate every
failure of an unbounded action space, but you can give the agent a loop that
observes each result and adapts — because "do anything" always includes "the
first attempt failed."
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
adr: convert all diagrams to Mermaid (sequenceDiagram, stateDiagram-v2, flowchart, erDiagram, graph)
Oikos
Agentic homelab operating system written in Go. Single binary (cmd/oikos),
Docker-deployed on mac-mini, with a standalone Nomos MCP agent gateway
(cmd/nomos). Manages the hubris Proxmox homelab autonomously — observes
state, classifies actions against policy, executes approved procedures over SSH,
learns from outcomes, and escalates when uncertain.
For agents running on enrolled clients: start with AGENTS.md. For client machines: see CLIENTS.md. For developers: see CONTRIBUTING.md.
Quick start
# Dev stack (postgres + api + scheduler + notifier)
docker compose --profile dev up -d
# Full stack (adds Nomos agent gateway)
docker compose --profile full up -d
# Build standalone binary
go build -o bin/oikos -tags timetzdata ./cmd/oikos
# Run all roles in one process (dev mode)
OIKOS_DATABASE_URL="postgres://oikos:oikos_dev@localhost:5432/oikos?sslmode=disable" \
go run ./cmd/oikos all
Architecture
┌──────────────────────────────────┐
│ mac-mini (Docker) │
│ │
Workstation ─── │ nomos (8092) ──MCP── api (8090) │
(mesh) │ MCP gateway REST + MCP │
│ │
│ scheduler ── notifier ── postgres │
│ (observe) (Matrix) (Timescale)│
└──────────────────────────────────┘
| Component | Port | Role |
|---|---|---|
oikos api |
8090 | REST API + MCP server (15 tools) |
oikos scheduler |
— | Probe runner, signal lifecycle, metrics |
oikos notifier |
— | Approval tokens, Matrix alerts |
nomos serve |
8092 | MCP client gateway, query routing |
Phases
| Phase | Status | Description |
|---|---|---|
| 1 — Ontology + DB | ✅ | TimescaleDB, migrations, seeds, blast_radius |
| 2 — API | ✅ | OpenAPI-first REST + MCP, auth, SSE, audit |
| 3 — Control loop | ✅ | Scheduler, actuator, learning, classifier, notifier |
| 4 — Nomos agent | ✅ | Standalone MCP client gateway, agent activity |
| 5 — Secrets | ✅ | Infisical backend + SOPS fallback, rotation runbooks |
| 6 — Deploy | ✅ | CI pipeline, cutover checklist, watchdog, rollback |
Full plan: plans/2026-07-06-consolidate-oikos-control-plane-onto-mac-mini.md.
Operations
API endpoints
curl http://localhost:8090/api/v1/entities?type=service # fleet
curl http://localhost:8090/api/v1/health # fleet health
curl http://localhost:8090/api/v1/agent-activity # agent log
Nomos queries
# Structured tool call
curl -X POST localhost:8092/query -H "Content-Type: application/json" \
-d '{"tool":"get_blast_radius","args":{"entity_id":"service:authentik"}}'
# Natural language
curl -X POST localhost:8092/query -H "Content-Type: application/json" \
-d '{"query":"what depends on authentik?"}'
CLI
oikos migrate # apply DB migrations
oikos seed # ingest ontology/inventory/policy seeds
oikos export # export DB state to YAML
oikos api # serve REST + MCP
oikos scheduler # run observe loop
oikos notifier # run notification loop
oikos all # all roles in one process
oikos secret list # enumerate SOPS secrets
oikos secret migrate # SOPS → Infisical
Repo layout
cmd/oikos/ Go entry point — single binary
cmd/nomos/ Nomos MCP client gateway
internal/ Go packages (httpapi, mcp, scheduler, actuator, learning,
notifier, policy, secrets, db, config, ontology, domain,
knowledge)
api/openapi.yaml API contract (OpenAPI 3.1)
migrations/ Forward-only SQL migrations (TimescaleDB)
seeds/ Bootstrap YAML (ontology, inventory, policy, knowledge)
compose/ Dockerfiles + Caddy config
scripts/ Deploy, watchdog, verification, rollback
nomos/ Nomos config, persona, skills
.agents/ Agent instruction files, shared conventions, skills
archive/ Historical reference (legacy wiki, plans, SOPS backups)
plans/ Design documents (active + done)
docs/adr/ Architecture decision records
For agents
See AGENTS.md for the full orientation. Quick reference:
- Source of truth: DB (runtime) then seeds (bootstrap). Old wiki is
archived at
archive/knowledge/— use MCPsearch_knowledgeinstead. - Mutations: classify against policy, request approval for
destructive/config_mutation - Secrets: Infisical (primary) or SOPS (fallback) — never hardcode
Related
- OIKOS.md — operating model, OODA loop, ontology
- CLIENTS.md — client onboarding guide
- CONTRIBUTING.md — developer guide
- plans/ — design documents and cutover checklist
- docs/adr/ — architecture decision records
Description
Languages
Go
53.1%
Svelte
25.7%
TypeScript
14%
Shell
3.8%
Python
1.7%
Other
1.5%