Manage watch expressions: add, remove, or list.
AI agents call debug_watch to retrieve information from Polybugger without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
While 'add' and 'remove' might suggest Write operations, watch expressions are debugging metadata that do not alter the target program's data or execution state. They are evaluation requests that return current values. The most prevalent use case is 'list' (Read). Even adding a watch is a transient debugging artifact, not a persistent data modification.
From the tool's definition Tool description states 'add, remove, or list' watch expressions. Watch expressions in debuggers are non-invasive monitoring constructs that track variable/expression values during debugging without modifying program state or executing arbitrary code.
Attacks that exploit this kind of access
Manage watch expressions: add, remove, or list. It is categorised as a Read tool in the Polybugger MCP Server, which means it retrieves data without modifying state.
Register the Polybugger MCP server in PolicyLayer and add a rule for debug_watch: 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 Polybugger. Nothing to install.
debug_watch 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 debug_watch 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 debug_watch. 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.
debug_watch is provided by the Polybugger MCP server (wilfoa/polybugger-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →