feat(lib): bundle MCP servers from .common/lib into factory/lib/

Copies source (no node_modules, dist, .tasks, .wiki, __pycache__) for:
- projects-meta-mcp v2.25.0 (TypeScript/Node)
- wiki-graph v0.3.1 (TypeScript/Node)
- interns-mcp v0.3.3 (Python/FastMCP)

.gitignore: exclude lib build artefacts (node_modules, dist, .venv, __pycache__, *.pyc)
bootstrap.ps1: add MCP build step — npm install+build for TS servers, venv+pip for Python

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

lib/interns-mcp/interns_mcp/interns/bulk_text_read.py

@@ -0,0 +1,46 @@
+"""Bulk text read intern — read N files and answer a focused question."""
+
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Any
+
+from .base import BlockedByPolicy, Intern, InternResponse
+
+
+class BulkTextRead(Intern):
+    id = "bulk_text_read"
+    description = "Read N files and answer a focused question. Returns concise summary."
+
+    def run(self, paths: list[str], question: str, **kwargs: Any) -> InternResponse | BlockedByPolicy:
+        contents: list[str] = []
+        for p in paths:
+            text = Path(p).read_text(encoding="utf-8", errors="replace")
+            contents.append(f"--- {p} ---\n{text}")
+
+        combined = "\n\n".join(contents)
+        prompt = f"Files:\n{combined}\n\nQuestion: {question}"
+
+        if self.client is None:
+            return InternResponse(text="No client configured", usage={})
+
+        extra_body = kwargs.get("extra_body", {})
+        resp = self.client.chat.completions.create(
+            model=self.model or "deepseek-v4-flash",
+            messages=[
+                {"role": "system", "content": self.system_prompt},
+                {"role": "user", "content": prompt},
+            ],
+            max_tokens=kwargs.get("max_tokens", self.max_tokens),
+            temperature=self.temperature,
+            extra_body=extra_body if extra_body else None,
+        )
+
+        choice = resp.choices[0]
+        return InternResponse(
+            text=choice.message.content or "",
+            usage={
+                "tokens_in": resp.usage.prompt_tokens if resp.usage else 0,
+                "tokens_out": resp.usage.completion_tokens if resp.usage else 0,
+            },
+        )