AI agents call svc_get_versions to retrieve information from Skill Vision Control (SVC) without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool performs a read-only retrieval of version metadata for a local skill. It queries existing state without modifying, deleting, executing code, or triggering external operations. This is a straightforward data retrieval task with minimal blast radius if misused by an AI agent.
From the tool's definition Tool name is 'svc_get_versions' with description 'Get all local versions of a skill' — this is a query operation that retrieves version information with no side effects.
Documented attack patterns abuse exactly the kind of access svc_get_versions gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Skill Vision Control (SVC), and nothing reaches the server without passing your rules. This is the rule we recommend for svc_get_versions:
{
"version": "1",
"default": "deny",
"tools": {
"svc_get_versions": {}
}
}svc_get_versions is read-only, so it stays allowed — but everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Get all local versions of a skill. It is categorised as a Read tool in the Skill Vision Control (SVC) MCP Server, which means it retrieves data without modifying state.
Register the Skill Vision Control (SVC) MCP server in PolicyLayer and add a rule for svc_get_versions: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Skill Vision Control (SVC). Nothing to install.
svc_get_versions is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the svc_get_versions rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for svc_get_versions. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
svc_get_versions is provided by the Skill Vision Control (SVC) MCP server (jane-xiaoer/skill-vision-control). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Skill Vision Control (SVC), add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
9 Skill Vision Control (SVC) tools catalogued and risk-classified — across an index of 43,000+ MCP servers.