Skills using-agent-relay

📦

using-agent-relay

Name: using-agent-relay
Author: AgentWorkforce

Content revision r2 High Risk ⚙️ External commands🌐 Network access

Coordinate Agents with Agent Relay

Multi-agent work can lose context when participants use inconsistent tools and status formats. This skill standardizes Relay registration, messaging, coordination, and worker delegation.

Supports: Claude Codex Code(CC)

⚠️ 38 Poor

Stop for confirmation before installing.

Review the plan and obtain explicit user consent before changing files.

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request

Review the Skillstore skill "using-agent-relay" from https://skillstore.io/skills/agentworkforce-using-agent-relay.md and its manifest at https://skillstore.io/api/skills/agentworkforce-using-agent-relay/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.

Your Agent should still show its plan and request any confirmation required by the security policy.

Agent-readable resources

Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.

Markdown detail GET /skills/agentworkforce-using-agent-relay.md Signed manifest GET /api/skills/agentworkforce-using-agent-relay/manifest Signed lockfile GET /api/skills/agentworkforce-using-agent-relay/lockfile

Test it

Using "using-agent-relay". Join the provided workspace as api-worker and acknowledge the lead.

Expected outcome:

Joined the workspace, registered as api-worker, checked the inbox, and sent the lead an acknowledgment for the assigned API scope.

Using "using-agent-relay". Find recent authentication blockers and update the team.

Expected outcome:

Found two authentication discussions, summarized the active blocker, contacted its owner, and posted the agreed next step in the working channel.

Using "using-agent-relay". Delegate a focused review and close the worker after acceptance.

Expected outcome:

Started a permitted review worker, tracked its response, shared the findings with the lead, and released the worker after acceptance.

Security Audit

High Risk

v9 • 7/23/2026 Open versioned report

All 77 external-command findings are false positives from Markdown backticks and fenced examples; the file contains no executable implementation. Four network findings are confirmed because agents are directed to hosted content or an external Relay gateway. Semantic review found unpinned remote instructions and credential examples that place secrets in command arguments.

Files scanned

232

Lines analyzed

Review items

False positives ignored

Confirmed security concerns (2)

High

SKILL.md:11-15 SKILL.md:195-200

Unpinned Remote Instructions for Agents

The skill tells agents and spawned workers to use a mutable hosted page. A compromised page could inject unaudited instructions into agent sessions.

Both locations explicitly direct an agent to use remote content whose version and integrity are not pinned in the audited file.

Medium

SKILL.md:205-211

Credentials Shown in Command Arguments

CLI examples place workspace keys and agent tokens in command arguments. Real values can remain in shell history and may be visible to local processes.

The examples repeatedly show workspace keys and tokens as flags, despite the later preference for environment variables.

Capability review items (4)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

Low

SKILL.md:3

Hardcoded URL

description: Use when you are a registered relay agent (a spawned worker, or a lead that called regi

The description directs agents to a hosted handoff URL, so following the skill can retrieve mutable external content.

Low

SKILL.md:15

Hardcoded URL

https://agentrelay.com/skill

The workflow explicitly says to use this hosted handoff before role selection, making external retrieval part of the documented flow.

Low

SKILL.md:200

Hardcoded URL

agent-relay node agent spawn claude --name Worker --task "Use https://agentrelay.com/skill and ACK o

The CLI example places the hosted URL in a spawned worker task, explicitly directing another agent to retrieve and use external instructions.

Low

SKILL.md:219

Hardcoded URL

RELAY_BASE_URL=https://gateway.relaycast.dev

The environment example configures gateway.relaycast.dev as the Relay endpoint, so operational use sends workspace traffic to an external service.

Risk Factors

⚙️ External commands (50)

🌐 Network access (5)

SKILL.md:3 SKILL.md:15 SKILL.md:21 SKILL.md:200 SKILL.md:219

Audited by: codex View Audit History →

Share & cite this report

Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.

Open versioned report

Copy report link

https://skillstore.io/skills/agentworkforce-using-agent-relay/audits/9?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown badge

[![Skillstore security assessment](https://skillstore.io/badges/skills/agentworkforce-using-agent-relay/security.svg)](https://skillstore.io/skills/agentworkforce-using-agent-relay?utm_source=security_passport_badge)

HTML badge

<a href="https://skillstore.io/skills/agentworkforce-using-agent-relay?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/agentworkforce-using-agent-relay/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

Embed card

<iframe src="https://skillstore.io/embed/skills/agentworkforce-using-agent-relay.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>

Academic citations (APA · BibTeX · CFF)

APA citation

AgentWorkforce. (2026). using-agent-relay security audit report (audit version 9) [Author version unspecified]. Skillstore. https://skillstore.io/skills/agentworkforce-using-agent-relay/audits/9

BibTeX citation

@techreport{agentworkforce-agentworkforce-using-agent-relay-2026,
  author = {AgentWorkforce},
  title = {using-agent-relay security audit report (audit version 9)},
  institution = {Skillstore},
  year = {2026},
  number = {9},
  url = {https://skillstore.io/skills/agentworkforce-using-agent-relay/audits/9},
  note = {Author version unspecified}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "using-agent-relay security audit report (audit version 9)"
version: "unspecified"
type: report
authors:
  - name: "AgentWorkforce"
date-released: "2026-07-23"
url: "https://skillstore.io/skills/agentworkforce-using-agent-relay/audits/9"
identifiers:
  - type: other
    value: "skillstore:agentworkforce-using-agent-relay:audit:9"
    description: "Skillstore immutable audit report identifier"

Skillstore Score

Why this score Evidence Confidence: Medium

Architecture

Maintainability

Content

Community

Spec Compliance

What You Can Build

Join a Relay Assignment

Coordinate Specialist Workers

Send targeted updates, organize channel discussions, delegate permitted work, and collect completion evidence.

Monitor Distributed Work

Review messages, threads, readers, reactions, and worker status across an active workspace.

Try These Prompts

Join and Acknowledge

Join workspace <workspace-key>, register as <agent-name>, check the inbox, and acknowledge <lead-name> before starting <scope>.

Review Team Activity

Review my Relay inbox and channel activity. Summarize priorities, identify blockers, and send concise status updates to the correct recipients.

Resolve a Shared Blocker

Search Relay messages for <topic>, inspect relevant threads, contact affected agents, and post a channel summary with owners and next steps.

Delegate and Close Work

Confirm my role permits delegation. Spawn a <provider> worker for <task>, monitor progress, collect evidence, report completion, and release the worker.

Best Practices

Confirm the workspace and registration state before calling participant-only tools.
Use direct messages for targeted coordination and channels for information relevant to the whole team.
Send acknowledgment, status, blocker, and completion updates with specific evidence.

Avoid

Do not use participant tools as a substitute for the outside orchestrator workflow.
Do not place real workspace keys or agent tokens in messages or command history.
Do not spawn workers unless the current role explicitly permits delegation.

Frequently Asked Questions

What must be configured before using this skill?

An Agent Relay MCP server or CLI must be available. You also need a workspace and a registered agent identity.

Does this skill start the local Relay broker?

No. It focuses on registered participants. Use the separate orchestrator workflow to start brokers and manage workers from outside Relay.

Can this skill create additional workers?

Yes, when the current role permits delegation and the Relay server exposes worker management tools.

Which AI clients are supported?

The marketplace metadata lists Claude, Codex, and Claude Code as supported tools.

How should credentials be handled?

Keep workspace keys and agent tokens in secure session or environment storage. Do not post them in messages or shared logs.

Does every Relay server expose the same tools?

No. Actions and structured result submission appear only when the server and spawned task enable them.

Developer Details

Author

AgentWorkforce

License

MIT

Skillstore revision

Version notice

The author did not declare a version.

Repository

https://github.com/AgentWorkforce/relay/tree/main/.claude/skills/using-agent-relay

Ref

31bc7611b504c197d3e993aef7943461c871616f

Maintenance freshness

7/26/2026

Usage

14 downloads · 331 views

File structure

📄 SKILL.md