XQE Agent Team
v1.0.0Orchestrate a dynamic multi-agent team where a lead agent (opus) plans and delegates tasks to specialized worker agents (sonnet) that communicate bidirection...
⭐ 0· 78·0 current·0 all-time
by@aa-rick
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description describe orchestrating an orchestrator + worker agents; SKILL.md contains only orchestration instructions (spawning subagents, passing messages, synthesizing results). There are no unrelated environment variables, binaries, or install steps requested.
Instruction Scope
Runtime instructions are focused on planning, spawning workers, relaying messages, and aggregating results. They reference platform session tools (sessions_spawn, sessions_send, subagents) which is expected for this purpose. The instructions do not ask to read local files, system configs, or unrelated env vars.
Install Mechanism
No install spec and no code files — instruction-only skill. Nothing is downloaded or written to disk by the skill package itself.
Credentials
The skill declares no required credentials or env vars (appropriate). One minor note: SKILL.md suggests overriding worker models with explicit model identifiers (e.g., "anthropic/claude-opus-4-6"); that implies the platform or runtime must have access to those models or keys, but the skill itself does not request or require credentials — this is a usage-time operational concern rather than a packaging inconsistency.
Persistence & Privilege
always is false and disable-model-invocation is not set — normal defaults. The skill does not request persistent installation, nor does it attempt to modify other skills or system-wide config.
Assessment
This skill is coherent for orchestrating multi-agent workflows, but you should still be cautious when using it: only provide sensitive data to an orchestrated team if you're comfortable multiple sub-agents will see it; be mindful of model overrides (they may require provider access or keys outside this skill); monitor spawned sessions and kill them when done to avoid runaway resource use; test workflows in a sandbox before running on production or confidential data. If you have policy or billing limits, restrict which models the skill may spawn and limit parallelism to control cost and blast radius.Like a lobster shell, security has layers — review code before you run it.
latestvk975dk6exabk5p3gjdqzy5sa0h83x9xp
License
MIT-0
Free to use, modify, and redistribute. No attribution required.