Configuration

memoryd reads configuration from ~/.memoryd/config.yaml on each team member's machine. Only the MongoDB connection string is required — everything else has sensible defaults.

Minimal setup (what most team members need)

mongodb_atlas_uri: "mongodb+srv://team-user:password@cluster0.mongodb.net/?retryWrites=true"
atlas_mode: true

That's it for most people. The connection string comes from whoever set up the shared cluster. atlas_mode: true turns on the full feature set.

Full reference

# Required — shared MongoDB Atlas connection string
mongodb_atlas_uri: "mongodb+srv://team:pass@cluster0.mongodb.net/?retryWrites=true"

# Local proxy port (default: 7432)
port: 7432

# MongoDB database name (default: "memoryd")
mongodb_database: "memoryd"

# Path to the local embedding model (auto-downloaded on first run)
model_path: "~/.memoryd/models/voyage-4-nano.gguf"

# Embedding dimensions — must match the Atlas vector index (default: 1024)
embedding_dim: 1024

# Number of knowledge items retrieved per query (default: 5)
retrieval_top_k: 5

# Maximum tokens of context returned per search (default: 2048)
retrieval_max_tokens: 2048

# Upstream LLM provider URL (default: https://api.anthropic.com)
upstream_anthropic_url: "https://api.anthropic.com"

# Enable Atlas hybrid search (default: false)
# Set to true when using a shared Atlas cluster
atlas_mode: false

# Quality maintenance settings
steward:
  interval_minutes: 60       # How often the quality sweep runs
  prune_threshold: 0.1       # Score below which low-value items are removed
  grace_period_hours: 24     # Minimum age before removal is considered
  decay_half_days: 90        # How quickly unused knowledge loses relevance
  merge_threshold: 0.88      # Similarity threshold for deduplication
  batch_size: 500            # Items processed per sweep

# Write pipeline noise filtering
pipeline:
  ingest_min_len: 80              # Responses shorter than this skip LLM synthesis
  content_score_pre_gate: 0.35    # Adaptive noise score threshold (0 = disabled)

Atlas mode

When atlas_mode: true, memoryd enables:

• Hybrid search — combines meaning-based and keyword-based retrieval for more accurate results
• Quality pre-filtering — search results are filtered by quality score at the database level
• Source-scoped search — filter results by knowledge source

This is the recommended setting for any team deployment using a shared Atlas cluster.

Tuning for your team

The defaults work well for most teams. If you need to adjust:

Scenario	What to change
Large team (10+ contributors)	Lower interval_minutes to 30, increase batch_size to 1000
Want to retain knowledge longer	Increase decay_half_days beyond 90
Seeing too many near-duplicates	Lower merge_threshold to 0.85
Want less aggressive cleanup	Lower prune_threshold to 0.05, increase grace_period_hours to 72
Heavy use of ingested sources	Increase batch_size to 2000, lower interval_minutes to 30
Too much noise getting stored	Lower content_score_pre_gate to 0.30 (more aggressive filtering)
Valuable short responses being dropped	Lower ingest_min_len to 40
Want to disable adaptive filtering	Set content_score_pre_gate to 0

Environment variables

The only environment variable relevant to memoryd is the one that connects your AI tool to the proxy:

export ANTHROPIC_BASE_URL=http://127.0.0.1:7432

This tells Claude Code (or any Anthropic-compatible tool) to route through memoryd. memoryd itself reads everything from config.yaml.