memlink-bridge 0.1.1

AI Memory Interchange Layer — language-neutral bridge for AI memory systems

memlink

A canonical interchange format for AI memories — like Pandoc, but for AI memory systems.

Built for developers building AI assistants, memory platforms, and agent frameworks.

---

Why memlink?

Different AI tools store memories differently. memlink lets them exchange memories without every project writing custom converters.

Without memlink — 10 memory formats need 45 converters. Every new format makes it worse:

Ombre → OpenClaw    Ombre → Mem0    Ombre → Zep
Mem0 → Zep          OpenClaw → Zep  ...

With memlink — each format writes one Reader + one Writer. Everything else is automatic:

         ┌── Reader → Canonical ──┬── Writer → OpenClaw
  Ombre ─┤                        ├── Writer → Mem0
         └── ...                   └── Writer → Any Format

n systems = 2n plugins. Not n² converters.

---

Quick Start

pip install memlink-bridge

# Convert Ombre → OpenClaw
memlink convert --from ombre --to openclaw \
  -s ~/.claude/ombre-buckets -T ./my-memories/

# Inspect any memory file
memlink inspect tests/fixtures/ombre_samples/dynamic/user/sample.md

# Validate data integrity
memlink validate -s tests/fixtures/ombre_samples --level schema

# Show installed formats
memlink formats

---

Supported Formats

Format	Read	Write	Status
Ombre Brain	✅	✅	Stable
OpenClaw	✅	✅	Stable
Generic Markdown	✅	—	Stable
Mem0	🚧	—	v0.2
Zep	🚧	—	v0.3

3 plugins, 7+ apps interoperable. Generic alone covers Obsidian, Logseq, Bear, iA Writer, and plain Markdown — every app that uses YAML frontmatter.

---

Architecture

  Ombre ──┐
Generic ──┼──→ Reader → Canonical Memory → Writer ──┬──→ OpenClaw
OpenClaw ─┘                                          └──→ Ombre

Each format implements three methods:

class FormatPlugin:
    def read(path) → ReadResult       # Format → Canonical
    def write(memories, path) → []    # Canonical → Format
    def validate(path) → [Issue]      # Integrity checks

Add a new format = write one plugin. Zero changes to core code.

---

Feature Compatibility

memlink is honest about what gets lost. Every conversion shows a Compatibility Report:

$ memlink convert --from ombre --to openclaw -s ombre/ -T openclaw/

  Read:     117 memories from ombre

  Compatibility Report:
    [ok] Preserved via metadata:
      Emotion fields (valence/arousal): 117 field values

  Warnings: 0
  Time:     0.23s

No silent data loss. No surprises.

---

Canonical Memory Schema

id: "project-alpha"
name: "Project Alpha Kickoff"
body: "..."
kind: dynamic
tags: [meeting, planning]
metadata: { ... }

See spec/canonical-v1.md for the full specification. JSON Schema also available.

---

Roadmap

Version	Focus
v0.2	Mem0 Reader, daily-notes roundtrip, --fail-on-loss
v0.3	Zep Reader, chat export readers (ChatGPT, Claude)
v0.4	memlink merge, memlink broadcast
v1.0	Stable Canonical Schema v1, stable Plugin API

---

What memlink is NOT

• ❌ Sync engine — v0 is export/import only
• ❌ Memory database — Works with files, not APIs
• ❌ Embedding store — No vector search
• ❌ Knowledge graph — No traversal or inference

---

Development

git clone https://github.com/velnori/memlink.git
cd memlink
pip install -e ".[dev]"

pytest tests/ -v          # 115 tests
ruff check memlink/       # Lint
mypy memlink/             # Type check

See CONTRIBUTING.md for how to add a new format.

---

License

MIT — see LICENSE.