AI agents use project_close to create or update resources in AudacityMCP — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your AudacityMCP environment.
This action modifies the state of the application and can result in loss of unsaved work, making it a Write-category operation (reversible state change) rather than Read. It is not Destructive because closing a project does not permanently delete files from disk—users can reopen projects. Severity is medium because misuse by an AI agent could cause loss of recent edits, but the underlying audio files remain intact.
From the tool's definition Tool name is 'project_close' and description states it will 'Close the current Audacity project.' Closing a project in audio editing software typically discards unsaved changes to the active working file.
Documented attack patterns abuse exactly the kind of access project_close gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and AudacityMCP, and nothing reaches the server without passing your rules. This is the rule we recommend for project_close:
{
"version": "1",
"default": "deny",
"tools": {
"project_close": {
"limits": [
{
"counter": "project_close_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}project_close 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.
Close the current Audacity project. It is categorised as a Write tool in the AudacityMCP MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Audacity MCP server in PolicyLayer and add a rule for project_close: 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 AudacityMCP. Nothing to install.
project_close 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 project_close 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 project_close. 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.
project_close is provided by the Audacity MCP server (xdarkzx/audacity-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 131 AudacityMCP tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
131 AudacityMCP tools catalogued and risk-classified — across an index of 42,500+ MCP servers.