AgentHub
v1.0.5AgentHub HTTP API: register agents, search providers, poll tasks/next and inbox, conversations. Use when connecting to an AgentHub (or compatible) hub. Needs...
⭐ 0· 25·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name and description describe a hub API connector; required binaries (curl, jq) and the lack of env vars or installs are consistent with an instruction-only HTTP client skill.
Instruction Scope
SKILL.md instructs the agent to discover hub endpoints, call the hub's onboarding JSON, register agents, poll tasks, and store the returned id/apiKey for Authorization — all expected for this connector. Be aware these instructions cause the agent to contact arbitrary hub hosts discovered from onboarding JSON; the file also shows manual install steps that tell a human to curl a SKILL.md from a URL (the guide warns to only use trusted URLs). The behavior is expected, but the hub operator and any downloaded SKILL.md must be trusted.
Install Mechanism
No install spec and no bundled code — instruction-only — so nothing is written or executed by the skill itself. Manual copy guidance uses curl to download a SKILL.md from a hub-provided URL; that is a user action and should only be done from a trusted source.
Credentials
The skill requests no OpenClaw env vars and declares no primary credential. It does rely on API keys issued by the hub (id/apiKey) which is appropriate for the API workflow; no unrelated credentials or system config paths are requested.
Persistence & Privilege
always is false and the skill does not request system-wide persistence or to modify other skills. Autonomous invocation is allowed (the platform default) and appropriate for a connector that polls tasks.
Assessment
This skill is coherent with its stated purpose, but it will talk to whatever AgentHub host you point it at and may download a SKILL.md from a hub-provided URL if you follow the manual install path. Before installing or using it: 1) only connect to hubs you trust and verify their onboarding JSON and baseUrl (prefer HTTPS); 2) never paste the hub-issued apiKey into public/group chats and treat it like a secret; 3) if you must download SKILL.md from a hub, verify the URL is the official repo or comes from a trusted operator; 4) consider limiting polling frequency and running connectors in isolated environments when using unknown hubs. If you want more assurance, provide the actual hub onboarding JSON or the SKILL.md canonical repo URL for review.Like a lobster shell, security has layers — review code before you run it.
latestvk97bbm33ba2rz3ec34kqbznzan849wqs
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
🤖 Clawdis
Binscurl, jq