Mache usage

Skill jamestexas/agents/skills/mache-usage

Curated library of Claude Code subagents and skills (23 agents, 20 skills) with an idempotent installer that symlinks into ~/.claude. Frontmatter-driven; spec-compliant with Anthropic and Gemini schemas.

Install

npx -y skills add jamestexas/agents --skill mache-usage

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Start or locate a running mache MCP server for a source path or .db file. Builds a .db from source if needed. Base skill used by diagram-gen-spec, diagram-gen-emergent, and review-prep. Returns port and session ID.

SKILL.md

3.9 KB, as published. Nobody here has run it

mache-usage — MCP Server Lifecycle

Ensure a mache MCP server is running for a given source and return the session info.

Arguments

$ARGUMENTS — <source-path-or-db> [port]

source: path to a directory of source files or a pre-built .db file
port: optional, default 7532

What To Do

Step 1: Parse arguments

SOURCE="<first argument>"
PORT="${2:-7532}"

Step 2: Check for existing server

curl -s --max-time 5 -X POST http://localhost:$PORT/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2024-11-05","capabilities":{},"clientInfo":{"name":"check","version":"1"}}}' \
  | python3 -c "import sys,json; d=json.load(sys.stdin); print('RUNNING' if 'result' in d and 'error' not in d else 'NOT_RUNNING')" 2>/dev/null || echo "NOT_RUNNING"

If RUNNING: skip to Step 4.

Step 3: Start the server

MACHE_BIN=$(which mache 2>/dev/null || echo "$HOME/go/bin/mache")
if [ ! -x "$MACHE_BIN" ]; then
  echo "status: unavailable — mache binary not found in PATH or ~/go/bin/"
  exit 0
fi

mache auto-detects schema from the source (Go files → go-schema.json, .db → serve directly). Do NOT pass --schema unless the user explicitly requests a specific schema.

$MACHE_BIN serve "$SOURCE" --http "localhost:$PORT" &
MACHE_PID=$!
for i in $(seq 1 10); do
  sleep 1
  curl -s --max-time 2 -X POST http://localhost:$PORT/mcp \
    -H "Content-Type: application/json" \
    -H "Accept: application/json, text/event-stream" \
    -d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2024-11-05","capabilities":{},"clientInfo":{"name":"check","version":"1"}}}' \
    | python3 -c "import sys,json; d=json.load(sys.stdin); exit(0) if 'result' in d else exit(1)" 2>/dev/null && echo "Server ready" && break
  if [ $i -eq 10 ]; then
    kill $MACHE_PID 2>/dev/null
    echo "status: unavailable — mache serve failed to start after 10 seconds"
    exit 0
  fi
done

Step 4: Initialize session and return info

SESSION=$(curl -s -X POST http://localhost:$PORT/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2024-11-05","capabilities":{},"clientInfo":{"name":"mache-usage","version":"1"}}}' \
  -D - 2>/dev/null | grep -i "^mcp-session-id:" | tr -d '\r' | awk -F': ' '{print $2}' | tr -d ' ')
if [ -z "$SESSION" ]; then
  echo "status: unavailable — session initialization failed"
  exit 0
fi
echo "status: available"
echo "port: $PORT"
echo "session: $SESSION"

Step 5: Graceful degradation

If the server fails to start after 10 seconds:

pkill -f "mache serve.*$PORT" 2>/dev/null

Report: status: unavailable — mache serve failed to start

Do NOT throw an error that blocks the caller. Return the status string and let the caller decide.

Helper: Call an MCP tool

Once you have port + session, call any mache tool:

mcp_call() {
  local PORT=$1 SESSION=$2 TOOL=$3 ARGS=$4
  curl -s -X POST http://localhost:$PORT/mcp \
    -H "Content-Type: application/json" \
    -H "Accept: application/json, text/event-stream" \
    -H "mcp-session-id: $SESSION" \
    -d "{\"jsonrpc\":\"2.0\",\"id\":2,\"method\":\"tools/call\",\"params\":{\"name\":\"$TOOL\",\"arguments\":$ARGS}}" \
  | python3 -c "import sys,json; d=json.load(sys.stdin); print(d['result']['content'][0]['text'])" 2>/dev/null
}

Usage: mcp_call $PORT $SESSION get_overview '{}'