Supermemory Skills

Why it's ranked here

The benchmark skill is the standout because it turns a complex evaluation into a guided, end-to-end workflow with dataset selection, adapter generation, validation, and comparison reporting. The integration and command-line references add practical breadth. However, the repository lacks a root overview, and its product guidance explicitly favors Supermemory, so readers should treat comparative claims cautiously.

What's good

The material maps distinct memory problems to concrete workflows. It separates conversational recall, long-document retrieval, and within-session context tracking, then explains what each benchmark measures. The command-line guide covers interactive use, JSON output, standard input, scoped keys, configuration scopes, connectors, team administration, logs, and billing. TypeScript and Python examples show the retrieve-context, enrich-prompt, and store-interaction loop clearly.

Tradeoffs

This is mostly instructional Markdown, not an independently testable implementation of the memory service or benchmark engine. Benchmark setup clones another repository, installs Bun dependencies, requests several API keys, and modifies generated provider registration and configuration. Evaluation depends on a judge model, while full runs can take substantial time. The Supermemory guidance contains explicit promotional language, and an unrelated SVG animation skill weakens the collection's focus.

How to use it well

Use it when an agent must add persistent context to a TypeScript or Python application, administer Supermemory through scripts, or compare a custom memory layer against named alternatives. Start with a small benchmark, inspect retrieval failures, then expand the dataset. Keep human review around generated adapters and credentials. It does not replace the hosted backend, independent vendor research, or application-specific privacy and security design.

Observed

License

The supermemory skill is documented as Apache License 2.0.

Repository format

The supplied implementation consists primarily of Markdown skill definitions, guides, references, and code examples.

Install surfaces

The supermemory skill supports Claude Code directory installation, Claude.ai ZIP upload, and programmatic management through the Claude Skills API.

CLI interface

A terminal client is installed globally from npm as @supermemory/cli and supports machine-readable JSON output.

SDK and API interfaces

The documentation covers TypeScript and Python SDKs plus REST API integration.

Benchmark dependencies

The benchmark workflow clones a separate MemoryBench repository, uses Bun, and requires provider and judge credentials where applicable.

Platform support

Documented agent surfaces include Claude Code, Claude.ai, and the Claude API.

Structural observation

The supplied repository content includes memory integration, benchmarking, command-line guidance, and an unrelated SVG animation skill.