super-memoriLocal-first hybrid memory skill for OpenClaw agents. Use when the agent needs to find, recall, search, or reuse past knowledge across episodic, semantic, pro...
Install via ClawdBot CLI:
clawdbot install ciklopentan/super-memoriGrade Fair — based on market validation, documentation quality, package completeness, maintenance status, and authenticity signals.
Potentially destructive shell commands in tool definitions
eval (Calls external URL not in known-safe list
http://127.0.0.1:6333AI Analysis
The skill's primary risk is the use of `eval()` and local network calls (127.0.0.1), which are flagged as unsafe patterns. However, these appear to be part of a local-first memory system and self-improvement loop, not for data exfiltration. There is no evidence of credential harvesting, unauthorized external servers, or obfuscated malicious behavior.
Audited Apr 17, 2026 · audit v1.0
Generated Oct 9, 2026
An OpenClaw agent handling multi-day support conversations for a SaaS product uses super-memori to remember prior tickets, resolved issues, and user preferences across sessions. When a customer asks 'what did we do about the billing sync issue last month', the skill retrieves both exact-match ticket IDs and semantically related past resolutions. This reduces repeated troubleshooting and gives support teams continuity without manual note-taking.
A researcher running a local Ubuntu workstation uses super-memori to store and query episodic and semantic memories of papers, experiments, and meeting notes. They can ask 'find in memory what we concluded about Qdrant vector dimensions' and get both lexical and semantic hits. The local-first design keeps sensitive research data off cloud services.
An SRE team deploys an OpenClaw agent that logs incident timelines, root causes, and remediation steps into super-memori as procedural memory. During new incidents, the agent recalls similar past failures and suggests known fixes, while the degraded-mode health checks prevent acting on stale or partial memory. This shortens mean time to resolution and builds an institutional memory of outages.
A user interacting with an OpenClaw agent in Russian and English relies on super-memori's semantic search to recall past conversations regardless of the language used originally. Queries like 'что мы делали' or 'remember this' route through hybrid retrieval to surface relevant episodic memories. The skill's weak-model decision matrix keeps responses safe even when semantic prerequisites are missing.
A regulated financial services firm uses super-memori on a local host to maintain audit trails of agent decisions and client interactions. Change-audit integration and memory-health guidance help ensure records are intact before any modification, and degraded-mode rules block risky writes when health checks return WARN or FAIL. This supports internal compliance reviews without sending data to third parties.
The super-memori skill is distributed freely as an open-source package for OpenClaw agents, with paid support contracts for organizations needing guaranteed SLAs, setup assistance on equipped hosts, and help tuning semantic retrieval prerequisites. Revenue comes from annual support subscriptions and priority bug fixes. This mirrors successful infrastructure OSS models but tailored to local-first agent memory.
A vendor packages super-memori with preconfigured Ubuntu hardware, local Qdrant, and embedding models as a turnkey appliance for enterprises that want agent memory without cloud exposure. Customers pay a hardware plus license fee, with optional maintenance for vector rebuilds and health monitoring. This targets regulated industries that value data sovereignty.
The core skill remains free, but advanced modules such as manual-review learning surfaces, memory-health dashboards, and pattern-mining tools are sold as premium add-ons through a skill marketplace. Teams can start with lexical-only mode and upgrade to hybrid semantic capabilities and audit integrations. Revenue is generated through per-seat or per-host premium licenses.
💬 Integration Tip
Run `./health-check.sh --json` before any query or write and respect the FAIL/WARN stop rules; ensure sentence-transformers, numpy, local Qdrant, and an embedding model are installed if you need semantic mode. Start with lexical-only queries via `./query-memory.sh --mode auto --json` while building vectors, and use `./memorize.sh` only for reusable lessons.
Scored Oct 9, 2026
Meta-skill for AI agent self-improvement. Analyzes runtime logs to detect error patterns, regressions, and inefficiencies, then generates structured improvem...
Stop waiting for prompts. Keep working.
Turn OpenClaw into a learning-loop agent with seeded workspace rules, skill promotion, reflective memory, and proactive maintenance.
Complete toolkit for creating autonomous AI agents and managing Discord channels for OpenClaw. Use when setting up multi-agent systems, creating new agents, or managing Discord channel organization.
Billions decentralized identity for agents. Link agents to human identities using Billions ERC-8004 and Attestation Registries. Verify and generate authentic...
AI agent for finance search agent tasks