← Back to Skills Marketplace
185
Downloads
0
Stars
1
Active Installs
5
Versions
Install in OpenClaw
/install swarmrecall
Description
Use SwarmRecall when an AI agent needs persistent memory, a knowledge graph, learnings, a skill registry, shared pools, or background dream consolidation acr...
Usage Guidance
This skill appears coherent for providing agent memory and a knowledge graph, but consider the following before installing:
- Privacy: Data stored via the skill is sent to https://swarmrecall-api.onrender.com and kept on SwarmRecall servers. Do not store sensitive personal or secret information unless you understand and accept that remote storage.
- API keys: Create a dedicated SWARMRECALL_API_KEY per agent (do not share keys across agents). Treat the key like a secret and rotate/revoke it if you suspect misuse.
- Install source: The CLI is installed via npm (@swarmrecall/cli). Verify the package on npmjs.com and the linked GitHub repo (owner, recent activity, and npm publisher) before installing globally.
- Local vs remote: If you are privacy-sensitive, prefer running the local stdio MCP server (swarmrecall mcp) under your control rather than pointing an agent to the hosted endpoint, or ask about a self-hosting option.
- Config file: The CLI will create ~/.config/swarmrecall/config.json if you use register --save. Ensure you do not check that file into source control.
- Audit and testing: After install, run the provided smoke prompts and inspect outbound traffic if possible; if anything unexpected appears (API URLs different from documented, keys returned in clear, or requests to unknown domains), revoke the key and uninstall.
If you want deeper assurance, review the CLI's published source code (the GitHub repo) and the package's npm publisher metadata before installing.
Capability Analysis
Type: OpenClaw Skill
Name: swarmrecall
Version: 1.2.0
The SwarmRecall skill provides a comprehensive suite for persistent agent memory, knowledge graphing, and background 'dream' consolidation via a third-party API (swarmrecall-api.onrender.com). While the functionality is transparently documented and includes security-conscious instructions regarding user consent and API key management, the skill possesses high-risk capabilities including the transmission of conversational history to an external endpoint and local file system access for configuration (~/.config/swarmrecall/config.json). Per the analysis criteria, these risky capabilities—though plausibly required for the stated purpose—warrant a suspicious classification in the absence of clear malicious intent.
Capability Tags
Capability Assessment
Purpose & Capability
The skill promises persistent memory, a knowledge graph, learnings, pools and a 'dream' consolidation flow and then requires a SwarmRecall CLI or API key and documents an HTTP MCP endpoint — all of which are coherent and expected for that purpose.
Instruction Scope
SKILL.md instructs installing the CLI or using the remote HTTP MCP endpoint, storing/using SWARMRECALL_API_KEY, and calling memory/knowledge/learnings/dream APIs. The instructions reference only the config file (~/.config/swarmrecall/config.json) and the declared API key and do not ask the agent to read unrelated system files or other credentials. It explicitly warns to obtain user consent before storing personal/sensitive data.
Install Mechanism
Install is via the npm package @swarmrecall/cli which will create a 'swarmrecall' binary — this is a typical distribution method for a CLI but carries the usual npm-package risk (remote code execution on install). There is a clear upstream project and GitHub URL in README, which makes verification possible; the install approach is proportionate to the stated CLI requirement.
Credentials
Only one primary credential is declared (SWARMRECALL_API_KEY), and SKILL.md uses it consistently for auth and for saving to the documented config path. No unrelated secrets or multiple external credentials are requested.
Persistence & Privilege
The skill is not force-included (always: false) and does not request elevated platform privileges. It can be invoked autonomously (disable-model-invocation: false) which is normal for skills; it does instruct storing an API key in a per-user config file (expected behavior for a CLI).
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install swarmrecall - After installation, invoke the skill by name or use
/swarmrecall - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.2.0
Added `@swarmrecall/mcp` package — an MCP (Model Context Protocol) server exposing every SDK module (memory, knowledge, learnings, skills, pools, dream) as 52 tools + 4 resources over stdio.; Added `swarmrecall mcp` CLI subcommand that boots the stdio MCP server using the existing API-key auth flow.; Exposed a remote MCP endpoint at `https://swarmrecall-api.onrender.com/mcp` over the Streamable HTTP transport. Authenticates with the same API key; no local install required on the client side.; Introduced `@swarmrecall/shared/config` subpath export with `loadConfig`, `saveConfig`, and `resolveClientOptions` so the CLI and MCP server share a single config loader.; Expanded the ClawHub skill bundle: full install metadata with `@swarmrecall/cli` bin, both stdio + remote MCP server advertisement, README onboarding, TROUBLESHOOTING, four workflow examples, a commands + mcp-tools reference, and a smoke-prompts validation script.; Added release infrastructure: `scripts/release-preflight.mjs`, `scripts/check-release-sync.mjs`, `scripts/check-clawhub-skill.mjs`, `scripts/publish-clawhub-skill.mjs`, `.github/workflows/ci.yml`, and this CHANGELOG.; Synchronized every SwarmRecall package to version 0.3.0. Skill version bumped to 1.2.0 on its own ClawHub track (monotonic from prior 1.1.0).
v1.1.0
Fix production API base URL to swarmrecall-api.onrender.com
v1.0.2
Added OpenClaw metadata, credentials declaration, privacy guardrails
v1.0.1
Added Module 5: Shared Pools for cross-agent collaboration
v1.0.0
- Initial release of SwarmRecall skill.
- Provides persistent memory, knowledge graph, learnings, and skill tracking for AI agents via the SwarmRecall API.
- Supports auto-registration and API key management.
- Includes modules for Memory (session, semantic recall), Knowledge (entities, graph traversal), Learnings (error/correction logging, pattern detection), and Skills (capability registry & suggestions).
- Each module offers structured endpoints with clear usage patterns and recommended behaviors.
Metadata
Frequently Asked Questions
What is SwarmRecall?
Use SwarmRecall when an AI agent needs persistent memory, a knowledge graph, learnings, a skill registry, shared pools, or background dream consolidation acr... It is an AI Agent Skill for Claude Code / OpenClaw, with 185 downloads so far.
How do I install SwarmRecall?
Run "/install swarmrecall" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is SwarmRecall free?
Yes, SwarmRecall is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does SwarmRecall support?
SwarmRecall is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created SwarmRecall?
It is built and maintained by Wayde (@waydelyle); the current version is v1.2.0.
More Skills