enhance_build
Run score_build and ask the configured LLM for bounded allowlisted improvements. Legacy calls are unchanged. Optional visualCritique uses the exact calibrated local vision receipt, explicit numeric targets, preview-only defaults, native Apply/Keep approval, CAS/readback, and compensating restore;...
This record as markdown: /tools/io-github-pantani-tdmcp/enhance-build.md
What enhance_build does on tdmcp — TouchDesigner MCP server
AI agents invoke enhance_build to trigger actions in tdmcp — TouchDesigner MCP server. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call: builds kicked off, notifications sent, workflows started.
Why enhance_build is rated High
While the tool includes approval gates ('native Apply/Keep approval', 'autoApply never bypasses approval'), it fundamentally executes code and makes modifications to a TouchDesigner project. This is Execute rather than Write because it runs an external process (score_build and LLM inference) and applies structural changes to a visual system.
From the tool's definition The tool 'enhance_build' explicitly runs code ('Run score_build'), triggers LLM processing ('ask the configured LLM'), and applies modifications to a visual system ('explicit numeric targets', 'Apply/Keep approval').
Attacks that exploit this kind of access
The rule that runs enhance_build safely
PolicyLayer is an MCP gateway: it sits between your AI agents and tdmcp — TouchDesigner MCP server, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For enhance_build, this is the rule to start with:
enhance_build stays usable, but rate-capped: a runaway agent can't fire it dozens of times a minute. Everything else on the server is denied unless you say otherwise.
The button opens the PolicyLayer dashboard: create your workspace, connect tdmcp — TouchDesigner MCP server, apply this rule, and every enhance_build call is checked against it from then on.
Questions about enhance_build
Run score_build and ask the configured LLM for bounded allowlisted improvements. Legacy calls are unchanged. Optional visualCritique uses the exact calibrated local vision receipt, explicit numeric targets, preview-only defaults, native Apply/Keep approval, CAS/readback, and compensating restore; autoApply never bypasses approval. It is categorised as a Execute tool in the tdmcp — TouchDesigner MCP server MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the tdmcp — TouchDesigner MCP server MCP server in PolicyLayer and add a rule for enhance_build: 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 tdmcp — TouchDesigner MCP server. Nothing to install.
enhance_build is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the enhance_build 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 enhance_build. 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.
enhance_build is provided by the tdmcp — TouchDesigner MCP server MCP server (@dpantani/tdmcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
More on tdmcp — TouchDesigner MCP server, and thousands of servers like it.
Across the catalogue