AI agents use publishPackage to create or update resources in Azure Devops — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Azure Devops environment.
Publishing a package to a feed creates new content in a package registry. This is a Write operation as it uploads/creates a package artifact. Severity is high because publishing malicious or incorrect packages to a feed can affect all consumers of that feed, potentially compromising downstream builds and deployments.
From the tool's definition Publish a package to a feed
Documented attack patterns abuse exactly the kind of access publishPackage gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Azure Devops, and nothing reaches the server without passing your rules. This is the rule we recommend for publishPackage:
{
"version": "1",
"default": "deny",
"tools": {
"publishPackage": {
"limits": [
{
"counter": "publishpackage_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}publishPackage stays usable, but capped — an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Publish a package to a feed. It is categorised as a Write tool in the Azure Devops MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Azure Devops MCP server in PolicyLayer and add a rule for publishPackage: 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 Azure Devops. Nothing to install.
publishPackage is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the publishPackage 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 publishPackage. 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.
publishPackage is provided by the Azure Devops MCP server (ryancardin15/azuredevops-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 97 Azure Devops tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
97 Azure Devops tools catalogued and risk-classified — across an index of 42,500+ MCP servers.