MCP Server for steering agent behavior
evaluate_tool_call
available_tools
: Array of tool definitions with schemasmessage_history
: Array of conversation messagesscore
: "correct"
, "incorrect_tool"
, "incorrect_parameter_names"
, or "incorrect_parameter_values"
reason
: Detailed explanation list.cursor/mcp.json
:
.cursor/rules/evaluate-tool-call.mdc
to always apply the evaluate_tool_call
function after any agent tool call.
%APPDATA%\amp\settings.json
~/.config/amp/settings.json
~/.config/amp/settings.json
.vscode/settings.json
). All settings use the amp.
prefix.
Then, add evaluation rules in your project directory as AGENT.md
:
uv
: brew install uv
(macOS) or follow uv installation guide~/Library/Application Support/Claude/claude_desktop_config.json
%APPDATA%\Claude\claude_desktop_config.json
Important: Replace3. Configure Personal Preferences (Recommended) To automatically evaluate tool usage, add this rule to your Claude Desktop preferences:/absolute/path/to/your/quotient-mcp/server.py
with the full absolute path to yourserver.py
file.
evaluate_tool_call
function after any agent tool call.