mnemo-cortex/robot.install at master · GuyMannDude/mnemo-cortex · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
{
  // ┌───────────────────────────────────────────────────────────────┐
  // │  THIS IS A SETTINGS FILE, NOT INSTALL INSTRUCTIONS.           │
  // │                                                               │
  // │  Looking for how to install Mnemo Cortex?  ->  ./INSTALL.md   │
  // │                                                               │
  // │  You do not need to edit this file to install Mnemo. The      │
  // │  defaults below already produce a working local install.      │
  // │  Change something here only if you want a different port,     │
  // │  data directory, or provider.                                 │
  // └───────────────────────────────────────────────────────────────┘
  //
  // Mnemo Cortex — robot.install manifest
  //
  // Non-interactive install recipe for AI agents and CI. Pair with
  // ./robot-install.sh. Single JSON object, // line comments allowed.
  //
  // The agent / operator fills in the values below, then runs:
  //   ./robot-install.sh
  // and parses the JSON object on stdout to confirm success.
  //
  // Sensible defaults are baked in — a manifest with only the
  // "service" block already produces a working local install. The
  // rest of the keys are optional and disabled / empty by default.

  "service": {
    "host": "127.0.0.1",
    "port": 50001,
    "data_dir": "~/.mnemo-cortex"
  },

  // Reasoning provider used for preflight checks and summarization.
  // Set api_key_env to a name the installer should pass through; the
  // installer reads that env var at install time and writes it into
  // the service unit so the running daemon picks it up.
  "reasoning": {
    "provider": "openrouter",
    "model": "google/gemini-2.5-flash",
    "api_base": "https://openrouter.ai/api/v1",
    "api_key_env": "OPENROUTER_API_KEY"
  },

  // Embedding provider — local-first by default. If you don't have
  // Ollama running, switch to a hosted provider (set provider to
  // openai/openrouter, model + api_key_env to match).
  "embedding": {
    "provider": "ollama",
    "model": "nomic-embed-text",
    "api_base": "http://localhost:11434"
  },

  // v3 decay thresholds. Override per-deployment. Topology decays
  // fast because IPs / ports / process state change often; relationship
  // facts (collaborator notes) decay slow.
  "decay": {
    "topology_warn_days": 30,
    "topology_stale_days": 90,
    "current_state_warn_days": 90,
    "relationship_warn_days": 180,
    "session_log_warn_days": 90
  },

  // systemd user-unit name. Set enabled=false to skip the systemd
  // step (for environments without systemd, e.g. macOS or containers).
  "systemd": {
    "enabled": true,
    "service_name": "mnemo-cortex"
  },

  // Smoke test exercises /health, then a save → recall round-trip.
  "smoke_test": {
    "enabled": true,
    "agent_id": "robot-install-smoke"
  },

  // Optional: seed the Phase 3 Facts table from a YAML of canonical truths so
  // exact lookups don't lose to stale fuzzy recall. Disabled by default. Point
  // "yaml" at your own file (see tools/seed-facts.example.yaml for the schema).
  // The watcher block can install a git post-commit hook and/or a nightly cron
  // that re-seed when the YAML changes — both off by default.
  "facts_seed": {
    "enabled": false,
    "yaml": "./tools/seed-facts.example.yaml",
    "watcher": {
      "post_commit_hook": false,
      "hook_repo": "",
      "nightly_cron": false,
      "cron_hour": 3,
      "cron_minute": 10
    }
  }
}