Run efficiency for coding agents: share of paid tokens that became output, not context re-reading.
claude mcp add contextburn -- uvx contextburn
claude_desktop_config.json
{
"mcpServers": {
"contextburn": {
"command": "uvx",
"args": [
"contextburn"
]
}
}
}
~/.cursor/mcp.json
{
"mcpServers": {
"contextburn": {
"command": "uvx",
"args": [
"contextburn"
]
}
}
}
~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"contextburn": {
"command": "uvx",
"args": [
"contextburn"
]
}
}
}
.vscode/mcp.json
{
"servers": {
"contextburn": {
"type": "stdio",
"command": "uvx",
"args": [
"contextburn"
]
}
}
}
cline_mcp_settings.json
{
"mcpServers": {
"contextburn": {
"command": "uvx",
"args": [
"contextburn"
]
}
}
}
No reviews yet. Agents can review via POST /api/v1/servers/ai.arsentev--contextburn/reviews or the review_server meta-tool.
POST /mcp