Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
ACP Orchestrator
v4.0.1Enhanced terminal AI agent orchestrator with parallel execution, health checks, and workflow presets.
⭐ 0· 638·4 current·4 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
The declared purpose (CLI orchestrator for agents) matches the instructions and manifest: the SKILL.md and manifest.json reference an 'acpx' CLI plus optional agent packages (opencode, @anthropic-ai/pi, kimi, etc.), which is coherent for an orchestrator. However, the registry summary above this bundle claims 'Required binaries: none' and 'Homepage: none' while manifest.json and SKILL.md explicitly list 'acpx' as a required binary and include a GitHub homepage/repository. This metadata mismatch is unexpected and should be confirmed.
Instruction Scope
SKILL.md contains only CLI usage for the acpx tool and examples of sessions_spawn(...) calls to spawn subagents (runtime='subagent', mode='run'). There are no instructions to read arbitrary files or environment variables in the included docs. Spawning subagents and passing tasks/flags is exactly the orchestrator's role, but it does mean the skill will cause execution of other agents/commands whose behavior depends on those agents — potential data forwarding or remote API interaction is possible via those agents.
Install Mechanism
The package is instruction-only (no code files) so the platform will not run installs automatically. However, manifest.json includes an 'install' section referencing npm packages ("acpx" and optional agent packages). This is not enforced by the registry metadata but indicates the author expects an npm-based install. Because there is no direct install spec executed by the platform, immediate install risk is low — nonetheless, if you follow the manifest's suggested npm installs, verify the packages and sources before running.
Credentials
The skill does not request environment variables or credentials in the registry metadata or SKILL.md. That is proportionate for a CLI orchestrator. However the manifest references optional external agent packages (e.g., '@anthropic-ai/pi') that in real use may require service credentials; the SKILL.md's sessions_spawn will pass tasks to subagents and could cause those agents to interact with external APIs. Make sure you do not inadvertently supply credentials to optional agents or allow the orchestrator to forward secrets.
Persistence & Privilege
always:false (no forced always-on) and default model invocation is allowed. The skill instructs spawning subagents (sessions_spawn), which is expected for an orchestrator but expands runtime activity. This is consistent with the stated purpose and does not itself request persistent system-wide privileges or modify other skills.
What to consider before installing
This skill is an orchestration wrapper and mostly consistent with that purpose, but there are several things to check before installing:
- Confirm the source: manifest.json references a GitHub repo (https://github.com/qriiz112/acpx-orchestrator) while the registry says 'Source: unknown' and 'Homepage: none' — verify the repository and author before trusting binaries.
- Verify the acpx binary/npm package: SKILL.md and manifest expect an 'acpx' CLI (and manifest suggests npm packages). If you plan to install those packages, inspect the npm package and GitHub repo for malicious code or unexpected install scripts.
- Be cautious with optional agents: the manifest lists optional external agents (Anthropic/pi, etc.). Those agents may require service credentials or send data externally. Do not provide secrets to those agents unless you trust them.
- Understand runtime behavior: the skill's examples use sessions_spawn to run subagents. Allowing the agent to spawn subagents means it can execute multiple agent processes and pass arbitrary tasks; limit what you let it run and review tasks it will forward.
- Metadata mismatches: the registry's top-level metadata (no required binaries/homepage) conflicts with files inside the package that declare requirements. Ask the publisher to clarify/correct metadata before installation.
If you cannot verify the package source or are uncomfortable with an orchestrator that spawns other agents, avoid installing or run it in an isolated environment (container/VM) and inspect any installed npm packages and their install scripts first.Like a lobster shell, security has layers — review code before you run it.
latestvk971cc8axqh0z54y2rw3b34b4n81yxzn
License
MIT-0
Free to use, modify, and redistribute. No attribution required.