@nodemint/projectmind - npm Package Compare versions

Comparing version 0.4.2 to 0.4.3

CHANGELOG.md

@@ -6,2 +6,18 @@ # Changelog
 
+## [0.4.3] - 2026-07-03
+
+### Added
+- **`projectmind setup --global`** — registers the MCP server once for every
+  future project, instead of per-repo `.mcp.json`. For Claude Code this shells
+  out to `claude mcp add --scope user` (the same mechanism codegraph and other
+  tools use for global registration) rather than hand-editing its internal
+  config; Cursor, Windsurf, and Gemini CLI get the same idempotent merge into
+  their global config files (`~/.cursor/mcp.json`,
+  `~/.codeium/windsurf/mcp_config.json`, `~/.gemini/settings.json`).
+
+Prompted by real dogfooding: without this, a fresh project needed
+`projectmind setup` run every time, unlike globally-registered MCP servers.
+Rules files (`CLAUDE.md` etc.) intentionally stay per-project — there's no
+sane "global" convention text — so `--global` only wires the server itself.
+
 ## [0.4.2] - 2026-07-03
@@ -8,0 +24,0 @@

package.json

 {
   "name": "@nodemint/projectmind",
-  "version": "0.4.2",
+  "version": "0.4.3",
   "mcpName": "io.github.Nodemint-dev/projectmind",
@@ -5,0 +5,0 @@ "publishConfig": {

README.md

@@ -243,2 +243,32 @@ # projectmind
 
+### Install once, use everywhere — global setup
+
+`projectmind setup` above is per-project. Run `projectmind setup --global`
+instead to register the MCP server **once**, so it's available in every
+project you open from then on — no per-repo `.mcp.json`, no re-running setup.
+Supported: Claude Code, Cursor, Windsurf, Gemini CLI.
+
+```bash
+projectmind setup --global
+```
+
+For Claude Code this shells out to its own `claude mcp add --scope user`
+(the same mechanism tools like codegraph use for global registration) rather
+than hand-editing its internal config file. For the others it merges into
+their global config (`~/.cursor/mcp.json`, `~/.codeium/windsurf/mcp_config.json`,
+`~/.gemini/settings.json`) the same idempotent way the per-project setup does.
+
+**Two things this can't do**, by design:
+- **Rules files stay per-project.** `CLAUDE.md` / `.cursorrules` / etc. are
+  checked-in project conventions — there's no sane "global" version of them,
+  so `--global` only registers the MCP *server*, not the nudge-text. Without
+  a project's `.projectmind/` map, `mind_digest` just returns an empty-but-valid
+  digest — harmless, but the agent won't be as strongly reminded to call it.
+- **A chat session already open won't see it.** Every MCP client (Claude Code,
+  Cursor, etc.) loads its server list once at session start — that's an MCP
+  client behavior, not something any server can override. Global scope means
+  every *new* session in every *future* project has it automatically; it does
+  not retroactively add it to a conversation already in progress. Restart the
+  agent once after running `--global`, and you're done for good.
+
 ## Where projectmind fits (and what it deliberately isn't)
@@ -245,0 +275,0 @@

src/cli/index.js

@@ -5,3 +5,3 @@ #!/usr/bin/env node
 import { installHook } from "../hooks/install.js";
-import { setupAgents, SUPPORTED_AGENTS } from "../setup/index.js";
+import { setupAgents, setupGlobalAgents, SUPPORTED_AGENTS, SUPPORTED_GLOBAL_AGENTS } from "../setup/index.js";
 import { watch } from "../watch/index.js";
@@ -18,2 +18,3 @@ import { savingsSummary, resolvePrice } from "../core/ledger.js";
   setup [--agent <name>]        Wire the MCP server + rules into agents (${SUPPORTED_AGENTS.join(", ")}, or all)
+  setup --global [--agent <name>]  Register the MCP server once, for every future project (${SUPPORTED_GLOBAL_AGENTS.join(", ")})
   digest                        Print the compact digest (what the agent reads first)
@@ -41,2 +42,3 @@ context [--files a,b] [--node id] [--term t] [--depth 1]   Task-scoped subgraph
   --agent <name>                (setup) target one agent; default all
+  --global                      (setup) register once for every future project, instead of just this repo
   --files <a,b,c>               (context) comma-separated files you're working on
@@ -60,2 +62,3 @@ --node <id>                   (context) seed the subgraph from this node
     else if (a === "--agent") opts.agent = argv[++i];
+    else if (a === "--global") opts.global = true;
     else if (a === "--files") opts.files = argv[++i];
@@ -115,2 +118,11 @@ else if (a === "--node") opts.node = argv[++i];
       const agents = opts.agent || "all";
+      if (opts.global) {
+        const results = setupGlobalAgents(agents);
+        out(`Registered projectmind globally (every future project will have it, no per-repo setup needed):`);
+        for (const x of results) {
+          out(`  ${x.status.padEnd(18)} ${x.file}${x.error ? ` — ${x.error}` : ""}`);
+        }
+        out("\nRestart your agent for this to take effect — including in any session already open.");
+        break;
+      }
       const results = setupAgents(r, agents);
@@ -117,0 +129,0 @@ out(`Wired projectmind into agents at ${r}:`);

src/mcp/server.js

@@ -182,3 +182,3 @@ #!/usr/bin/env node
   const server = new Server(
-    { name: "projectmind", version: "0.4.2" },
+    { name: "projectmind", version: "0.4.3" },
     { capabilities: { tools: {} } }
@@ -185,0 +185,0 @@ );

src/setup/index.js

@@ -5,4 +5,13 @@ // Multi-agent wiring. Writes the MCP server config and a workflow rules block
 // rather than destroy it. Rules files get a marked, idempotent block.
+//
+// Two scopes:
+//  - project (default): writes .mcp.json etc. in the current repo only.
+//  - global: registers projectmind once, for every future project — the same
+//    "install once, works everywhere" model as codegraph's user-scoped MCP
+//    entry. No rules-file text is global (CLAUDE.md/AGENTS.md are inherently
+//    per-repo), only the MCP server registration.
 import fs from "node:fs";
+import os from "node:os";
 import path from "node:path";
+import { execFileSync } from "node:child_process";
 
@@ -60,4 +69,5 @@ // `mcp` is a subcommand of the main bin, so plain npx works (same invocation
 
-function mergeJsonConfig(r, spec) {
-  const file = path.join(r, spec.file);
+// file: absolute path. label: what to report back (may differ from `file`,
+// e.g. "~/.cursor/mcp.json" for readability).
+function mergeJsonConfig(file, key, label = file) {
   let data = {};
@@ -70,11 +80,11 @@ if (fs.existsSync(file)) {
       try { fs.copyFileSync(file, backup); } catch { /* ignore */ }
-      return { file: spec.file, status: "skipped-unparseable", backup };
+      return { file: label, status: "skipped-unparseable", backup };
     }
   }
-  data[spec.key] = data[spec.key] || {};
-  if (data[spec.key].projectmind) return { file: spec.file, status: "already" };
-  data[spec.key].projectmind = { ...MCP_ENTRY };
+  data[key] = data[key] || {};
+  if (data[key].projectmind) return { file: label, status: "already" };
+  data[key].projectmind = { ...MCP_ENTRY };
   fs.mkdirSync(path.dirname(file), { recursive: true });
   fs.writeFileSync(file, JSON.stringify(data, null, 2) + "\n");
-  return { file: spec.file, status: "written" };
+  return { file: label, status: "written" };
 }
@@ -100,3 +110,3 @@
     if (!a) continue;
-    if (a.json) results.push({ agent: key, label: a.label, ...mergeJsonConfig(r, a.json) });
+    if (a.json) results.push({ agent: key, label: a.label, ...mergeJsonConfig(path.join(r, a.json.file), a.json.key, a.json.file) });
     if (a.rules) results.push({ agent: key, label: a.label, ...appendRulesBlock(r, a.rules) });
@@ -106,1 +116,49 @@ }
 }
+
+// ---------------------------------------------------------------------------
+// Global (user) scope — register once, available in every future project
+// without a per-project .mcp.json. Only the MCP server entry is global; rules
+// files (CLAUDE.md, AGENTS.md, ...) are inherently per-repo and untouched.
+// ---------------------------------------------------------------------------
+export const SUPPORTED_GLOBAL_AGENTS = ["claude", "cursor", "windsurf", "gemini"];
+
+function globalJsonPath(agent, homeDir) {
+  switch (agent) {
+    case "cursor": return path.join(homeDir, ".cursor", "mcp.json");
+    case "windsurf": return path.join(homeDir, ".codeium", "windsurf", "mcp_config.json");
+    case "gemini": return path.join(homeDir, ".gemini", "settings.json");
+    default: return null;
+  }
+}
+
+// Claude Code has no documented global config file to hand-edit safely, but
+// ships a first-class CLI verb for exactly this (`claude mcp add --scope
+// user`) — shell out to it rather than guess an internal file format.
+function addClaudeUserScope(execFile = execFileSync) {
+  const label = "claude mcp (user scope)";
+  try {
+    execFile("claude", ["mcp", "add", "projectmind", "--scope", "user", "--", "npx", "-y", "@nodemint/projectmind", "mcp"], { stdio: "pipe" });
+    return { file: label, status: "written" };
+  } catch (err) {
+    const msg = String(err?.stderr || err?.message || "");
+    if (/already exists/i.test(msg)) return { file: label, status: "already" };
+    return { file: label, status: "failed", error: (msg.split("\n").find(Boolean) || "is the `claude` CLI installed and on PATH?").trim() };
+  }
+}
+
+// homeDir is overridable for tests; defaults to the real home directory.
+export function setupGlobalAgents(agents = "all", { homeDir = os.homedir(), execFile = execFileSync } = {}) {
+  const keys = agents === "all" || !agents ? SUPPORTED_GLOBAL_AGENTS : [].concat(agents).filter((k) => SUPPORTED_GLOBAL_AGENTS.includes(k));
+  const results = [];
+  for (const key of keys) {
+    if (key === "claude") {
+      results.push({ agent: "claude", label: "Claude Code", ...addClaudeUserScope(execFile) });
+      continue;
+    }
+    const file = globalJsonPath(key, homeDir);
+    if (!file) continue;
+    const label = key === "cursor" ? "~/.cursor/mcp.json" : key === "windsurf" ? "~/.codeium/windsurf/mcp_config.json" : "~/.gemini/settings.json";
+    results.push({ agent: key, label: AGENTS[key]?.label || key, ...mergeJsonConfig(file, "mcpServers", label) });
+  }
+  return results;
+}

New alerts

Shell access

Supply chain risk

This module accesses the system shell. Accessing the system shell increases the risk of executing arbitrary code.

AI-detected potential code anomaly

Supply chain risk

AI has identified unusual behaviors that may pose a security risk.

Long strings

Supply chain risk

Contains long string literals, which may be a sign of obfuscated or packed code.

URL strings

Supply chain risk

Package contains fragments of external URLs or IP addresses, which the package may be accessing at runtime.

Fixed alerts

AI-detected potential code anomaly

Supply chain risk

AI has identified unusual behaviors that may pose a security risk.

Long strings

Supply chain risk

Contains long string literals, which may be a sign of obfuscated or packed code.

URL strings

Supply chain risk

Package contains fragments of external URLs or IP addresses, which the package may be accessing at runtime.

Improved metrics

Total package byte prevSize

103937

6.5%

Lines of code

1700

4.04%

Number of lines in readme file

413

7.83%

Worsened metrics

Number of medium supply chain risk alerts

2

100%

No dependency changes