memorix/memorix.example.yml at main · AVIDS2/memorix · 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
# ============================================================
# memorix.yml — Memorix Configuration
# ============================================================
#
# Two files, two roles:
#   memorix.yml  → behavior settings (this file)
#   .env         → secrets only (API keys) — see .env.example
#
# Place this file at:
#   - Project root (./memorix.yml)  — project-specific settings
#   - User home (~/.memorix/memorix.yml) — global defaults (advanced)
#
# Priority: env > .env > project memorix.yml > user memorix.yml > defaults
#
# Docs: https://github.com/AVIDS2/memorix

# ── LLM Configuration ────────────────────────────────────────
# Used for smart dedup, fact extraction, and memory formation.
# Optional — Memorix works without LLM (free heuristic mode).
# API key goes in .env as MEMORIX_LLM_API_KEY
llm:
  provider: openai          # openai, anthropic, openrouter, or custom endpoint
  model: gpt-4o-mini
  # baseUrl: https://api.openai.com/v1

# ── Embedding / Vector Search ──────────────────────────────
# Enables semantic search alongside BM25 fulltext.
# API key goes in .env as MEMORIX_EMBEDDING_API_KEY
embedding:
  provider: off             # off (default), api, fastembed, transformers, auto
  # model: text-embedding-3-small
  # dimensions: 1024        # Override auto-detected dimensions

# ── Git-Memory Pipeline ──────────────────────────────────────
# Automatically capture engineering knowledge from git commits.
# Every commit becomes a searchable, immutable memory.
git:
  autoHook: false           # Auto-install post-commit hook on first serve
  ingestOnCommit: true      # Ingest HEAD commit via post-commit hook
  maxDiffSize: 500          # Max diff chars included in memory
  skipMergeCommits: true    # Skip merge commits (usually noise)
  # excludePatterns:        # File glob patterns — commits touching ONLY these are skipped
  #   - "*.lock"
  #   - "dist/**"
  # noiseKeywords:          # Regex patterns — commits matching these subjects are skipped
  #   - "^BOT:"             # Skip bot-generated commits
  #   - "auto-deploy"       # Skip CI deploy commits

# ── Behavior Settings ────────────────────────────────────────
behavior:
  sessionInject: minimal    # full | minimal | silent
  syncAdvisory: true        # Show cross-agent sync hint on first search
  autoCleanup: true         # Auto-archive expired memories on startup
  formationMode: active     # active | shadow | fallback

# ── Server Configuration ─────────────────────────────────────
server:
  transport: stdio          # stdio (default) or http
  # port: 37850             # HTTP port (only for http transport)
  dashboard: true           # Enable Web Dashboard
  dashboardPort: 3210       # Dashboard port

# ── Team Collaboration ───────────────────────────────────────
# team:
#   enabled: false
#   workspaceCollection: workspace_memory

# ── MCP Server Aggregation (Hub Mode) ────────────────────────
# When Memorix acts as a central hub, it can proxy other MCP servers.
# mcpServers:
#   filesystem:
#     type: stdio
#     command: npx
#     args: ["-y", "@modelcontextprotocol/server-filesystem", "."]