v0.1.1: Rewrite with curl-based HTTP (Cloudflare bypass), S3 gzip content fetching, correct API field names

- Switch urllib to curl subprocess to bypass Cloudflare Error 1010
- Add _fetch_s3_json() for gzipped content from signed S3 URLs  
- Fix list endpoint key: data_file_list (not data)
- Fix detail endpoint title field: file_name (not filename)
- Fix token key: support both token and accessToken in config
- Add _format_transcript() with speaker grouping
- plaud_get_transcript: fetch from content_list S3 URLs (transaction_polish → transaction)
- plaud_get_summary: fetch outline from S3
- plaud_get_notes: fetch all non-transcript content_list items from S3
- plaud_get_recording_detail: derive has_transcript from content_types list

server/plaud_mcp.py

@@ -1,338 +1 @@
-#!/usr/bin/env python3
-"""
-Plaud MCP Server for MPM
-Connects to the Plaud API and exposes tools for listing, searching,
-and reading transcripts, summaries, and notes from Plaud recordings.
-
-Credentials: reads token from ~/.plaud/config.json (written by plaud-connector
-import-token command). Falls back to PLAUD_TOKEN env var.
-"""
-
-import json
-import os
-import urllib.request
-import urllib.error
-import urllib.parse
-from datetime import datetime, timezone
-from pathlib import Path
-from typing import Optional
-
-from mcp.server.fastmcp import FastMCP
-
-# ── Configuration ─────────────────────────────────────────────────────────────
-# US region: api.plaud.ai  |  EU region: api.plaud.eu
-PLAUD_REGION = os.environ.get("PLAUD_REGION", "us")
-BASE_URL = "https://api.plaud.eu" if PLAUD_REGION == "eu" else "https://api.plaud.ai"
-CONFIG_PATH = Path.home() / ".plaud" / "config.json"
-
-
-# ── Token resolution ──────────────────────────────────────────────────────────
-
-def _load_token() -> Optional[str]:
-    """
-    Resolve the Plaud bearer token. Priority:
-      1. PLAUD_TOKEN env var
-      2. ~/.plaud/config.json  → token.token  (plaud-connector format)
-      3. ~/.plaud/config.json  → token  (bare string fallback)
-    """
-    env_token = os.environ.get("PLAUD_TOKEN", "").strip()
-    if env_token:
-        return env_token
-
-    if CONFIG_PATH.exists():
-        try:
-            raw = CONFIG_PATH.read_text()
-            data = json.loads(raw)
-            # plaud-connector / plaud-toolkit format: { "token": { "token": "...", "expiresAt": ... } }
-            token_block = data.get("token")
-            if isinstance(token_block, dict):
-                t = token_block.get("token", "").strip()
-                if t:
-                    return t
-            # bare string fallback
-            if isinstance(token_block, str) and token_block.strip():
-                return token_block.strip()
-        except Exception:
-            pass
-
-    return None
-
-
-def _get_token() -> str:
-    token = _load_token()
-    if not token:
-        raise RuntimeError(
-            "No Plaud token found. Run:\n"
-            "  npx tsx ~/Developer/plaud-connector/packages/cli/bin/plaud.ts import-token app \"<jwt>\"\n"
-            "or set the PLAUD_TOKEN environment variable."
-        )
-    return token
-
-
-# ── HTTP helpers ──────────────────────────────────────────────────────────────
-
-def _api_get(path: str, params: Optional[dict] = None) -> dict:
-    """GET request to the Plaud API. Returns parsed JSON."""
-    token = _get_token()
-    url = f"{BASE_URL}{path}"
-    if params:
-        query = "&".join(f"{k}={urllib.parse.quote(str(v))}" for k, v in params.items() if v is not None)
-        if query:
-            url = f"{url}?{query}"
-
-    req = urllib.request.Request(url)
-    req.add_header("Authorization", f"Bearer {token}")
-    req.add_header("Content-Type", "application/json")
-    req.add_header("Accept", "application/json")
-    req.add_header("User-Agent", "plaud-mpm-cowork/0.1.0")
-
-    try:
-        with urllib.request.urlopen(req, timeout=30) as resp:
-            return json.loads(resp.read().decode("utf-8"))
-    except urllib.error.HTTPError as e:
-        body = e.read().decode("utf-8", errors="replace")
-        raise RuntimeError(f"Plaud API error {e.code}: {body[:300]}")
-    except urllib.error.URLError as e:
-        raise RuntimeError(f"Network error reaching Plaud API: {e.reason}")
-
-
-# ── Data helpers ──────────────────────────────────────────────────────────────
-
-def _fmt_duration(ms: int) -> str:
-    mins = ms // 60000
-    if mins < 60:
-        return f"{mins}m"
-    return f"{mins // 60}h {mins % 60}m"
-
-
-def _fmt_date(ts_ms: int) -> str:
-    try:
-        dt = datetime.fromtimestamp(ts_ms / 1000, tz=timezone.utc)
-        return dt.strftime("%Y-%m-%d %H:%M UTC")
-    except Exception:
-        return str(ts_ms)
-
-
-def _summarise_recording(r: dict) -> dict:
-    """Return a compact dict suitable for listing."""
-    return {
-        "id": r.get("id", ""),
-        "title": r.get("filename") or r.get("title") or "(untitled)",
-        "date": _fmt_date(r.get("start_time", 0)),
-        "duration": _fmt_duration(r.get("duration", 0)),
-        "has_transcript": bool(r.get("is_trans")),
-        "has_summary": bool(r.get("auto_sum_note") or r.get("is_sum")),
-    }
-
-
-# ── MCP server ────────────────────────────────────────────────────────────────
-mcp = FastMCP("plaud-mpm")
-
-
-@mcp.tool()
-def plaud_list_recordings(
-    limit: int = 50,
-    only_with_transcript: bool = False,
-) -> str:
-    """
-    List Plaud recordings newest-first.
-
-    Args:
-        limit: Max number to return (default 50, max 200).
-        only_with_transcript: If true, only return recordings that have a transcript.
-    """
-    resp = _api_get("/file/simple/web")
-    recordings = resp.get("data", resp) if isinstance(resp, dict) else resp
-    if not isinstance(recordings, list):
-        recordings = []
-
-    if only_with_transcript:
-        recordings = [r for r in recordings if r.get("is_trans")]
-
-    recordings = recordings[:min(limit, 200)]
-    result = [_summarise_recording(r) for r in recordings]
-    return json.dumps(result, indent=2)
-
-
-@mcp.tool()
-def plaud_search_recordings(
-    title_contains: Optional[str] = None,
-    start_date: Optional[str] = None,
-    end_date: Optional[str] = None,
-    only_with_transcript: bool = False,
-    max_results: int = 20,
-) -> str:
-    """
-    Search recordings by title and/or date range.
-
-    Args:
-        title_contains: Substring to match in the recording title (case-insensitive).
-        start_date: ISO date string, e.g. '2026-04-01'. Only recordings on or after this date.
-        end_date: ISO date string, e.g. '2026-05-01'. Only recordings on or before this date.
-        only_with_transcript: If true, only return recordings that have a transcript.
-        max_results: Max number of results to return (default 20).
-    """
-    resp = _api_get("/file/simple/web")
-    recordings = resp.get("data", resp) if isinstance(resp, dict) else resp
-    if not isinstance(recordings, list):
-        recordings = []
-
-    def matches(r: dict) -> bool:
-        if only_with_transcript and not r.get("is_trans"):
-            return False
-        title = (r.get("filename") or r.get("title") or "").lower()
-        if title_contains and title_contains.lower() not in title:
-            return False
-        if start_date or end_date:
-            ts_ms = r.get("start_time", 0)
-            rec_date = datetime.fromtimestamp(ts_ms / 1000, tz=timezone.utc).date()
-            if start_date:
-                if rec_date < datetime.fromisoformat(start_date).date():
-                    return False
-            if end_date:
-                if rec_date > datetime.fromisoformat(end_date).date():
-                    return False
-        return True
-
-    matched = [r for r in recordings if matches(r)][:max_results]
-    return json.dumps([_summarise_recording(r) for r in matched], indent=2)
-
-
-@mcp.tool()
-def plaud_get_transcript(recording_id: str) -> str:
-    """
-    Get the full transcript for a Plaud recording.
-
-    Args:
-        recording_id: The recording ID (from plaud_list_recordings).
-    """
-    resp = _api_get(f"/file/detail/{recording_id}")
-    data = resp.get("data", resp) if isinstance(resp, dict) else resp
-
-    title = data.get("filename") or data.get("title") or "(untitled)"
-    transcript = data.get("transcript") or ""
-
-    # Some recordings store transcript in source_list segments
-    if not transcript:
-        source_list = data.get("source_list") or []
-        if source_list:
-            transcript = "\n".join(
-                f"[{s.get('speaker', '')}] {s.get('content', '')}" if s.get("speaker")
-                else s.get("content", "")
-                for s in source_list
-            )
-
-    return json.dumps({
-        "id": recording_id,
-        "title": title,
-        "date": _fmt_date(data.get("start_time", 0)),
-        "duration": _fmt_duration(data.get("duration", 0)),
-        "transcript": transcript or "No transcript available for this recording.",
-    }, indent=2)
-
-
-@mcp.tool()
-def plaud_get_summary(recording_id: str) -> str:
-    """
-    Get the AI-generated summary for a Plaud recording.
-
-    Args:
-        recording_id: The recording ID (from plaud_list_recordings).
-    """
-    resp = _api_get(f"/file/detail/{recording_id}")
-    data = resp.get("data", resp) if isinstance(resp, dict) else resp
-
-    title = data.get("filename") or data.get("title") or "(untitled)"
-    summary = data.get("auto_sum_note") or ""
-
-    return json.dumps({
-        "id": recording_id,
-        "title": title,
-        "date": _fmt_date(data.get("start_time", 0)),
-        "summary": summary or "No AI summary available for this recording.",
-    }, indent=2)
-
-
-@mcp.tool()
-def plaud_get_notes(recording_id: str) -> str:
-    """
-    Get AI-generated notes and action items for a Plaud recording.
-    Returns all note entries from the content_list (action items, key points, etc.).
-
-    Args:
-        recording_id: The recording ID (from plaud_list_recordings).
-    """
-    resp = _api_get(f"/file/detail/{recording_id}")
-    data = resp.get("data", resp) if isinstance(resp, dict) else resp
-
-    title = data.get("filename") or data.get("title") or "(untitled)"
-    content_list = data.get("content_list") or []
-
-    notes = []
-    for item in content_list:
-        note_entry = {
-            "type": item.get("data_type") or item.get("template_type") or "note",
-            "title": item.get("template_name") or item.get("tab_name") or "",
-            "content": item.get("note") or item.get("content") or "",
-        }
-        if note_entry["content"]:
-            notes.append(note_entry)
-
-    return json.dumps({
-        "id": recording_id,
-        "title": title,
-        "date": _fmt_date(data.get("start_time", 0)),
-        "notes": notes if notes else [{"type": "none", "content": "No notes available for this recording."}],
-    }, indent=2)
-
-
-@mcp.tool()
-def plaud_get_recording_detail(recording_id: str) -> str:
-    """
-    Get full metadata for a Plaud recording including title, date, duration,
-    transcript availability, summary availability, and note count.
-
-    Args:
-        recording_id: The recording ID (from plaud_list_recordings).
-    """
-    resp = _api_get(f"/file/detail/{recording_id}")
-    data = resp.get("data", resp) if isinstance(resp, dict) else resp
-
-    return json.dumps({
-        "id": recording_id,
-        "title": data.get("filename") or data.get("title") or "(untitled)",
-        "date": _fmt_date(data.get("start_time", 0)),
-        "duration": _fmt_duration(data.get("duration", 0)),
-        "has_transcript": bool(data.get("is_trans")),
-        "has_summary": bool(data.get("auto_sum_note") or data.get("is_sum")),
-        "note_count": len(data.get("content_list") or []),
-        "language": data.get("language") or "unknown",
-        "device": data.get("device_name") or data.get("device") or "unknown",
-        "tags": data.get("filetag") or [],
-    }, indent=2)
-
-
-@mcp.tool()
-def plaud_user_info() -> str:
-    """Get the current Plaud account information."""
-    try:
-        resp = _api_get("/user/info")
-        data = resp.get("data", resp) if isinstance(resp, dict) else resp
-        return json.dumps({
-            "email": data.get("email") or data.get("username") or "unknown",
-            "name": data.get("name") or data.get("display_name") or "",
-            "region": PLAUD_REGION,
-            "api_base": BASE_URL,
-            "token_loaded": bool(_load_token()),
-        }, indent=2)
-    except Exception as e:
-        return json.dumps({
-            "region": PLAUD_REGION,
-            "api_base": BASE_URL,
-            "token_loaded": bool(_load_token()),
-            "note": str(e),
-        }, indent=2)
-
-
-if __name__ == "__main__":
-    mcp.run()
+IyEvdXNyL2Jpbi9lbnYgcHl0aG9uMwoiIiIKUGxhdWQgTUNQIFNlcnZlciBmb3IgTVBNCkNvbm5lY3RzIHRvIHRoZSBQbGF1ZCBBUEkgYW5kIGV4cG9zZXMgdG9vbHMgZm9yIGxpc3RpbmcsIHNlYXJjaGluZywKYW5kIHJlYWRpbmcgdHJhbnNjcmlwdHMsIHN1bW1hcmllcywgYW5kIG5vdGVzIGZyb20gUGxhdWQgcmVjb3JkaW5ncy4KCkNyZWRlbnRpYWxzOiByZWFkcyB0b2tlbiBmcm9tIH4vLnBsYXVkL2NvbmZpZy5qc29uLgoiIiIKCmltcG9ydCBnemlwCmltcG9ydCBqc29uCmltcG9ydCBvcwppbXBvcnQgc3VicHJvY2VzcwppbXBvcnQgdXJsbGliLnBhcnNlCmZyb20gZGF0ZXRpbWUgaW1wb3J0IGRhdGV0aW1lLCB0aW1lem9uZQpmcm9tIHBhdGhsaWIgaW1wb3J0IFBhdGgKZnJvbSB0eXBpbmcgaW1wb3J0IE9wdGlvbmFsCgpmcm9tIG1jcC5zZXJ2ZXIuZmFzdG1jcCBpbXBvcnQgRmFzdE1DUAoKIyDilIDilIAgQ29uZmlndXJhdGlvbiDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIAKUExBVURfUkVHSU9OID0gb3MuZW52aXJvbi5nZXQoIlBMQVVEX1JFR0lPTiIsICJ1cyIpCkJBU0VfVVJMID0gImh0dHBzOi8vYXBpLnBsYXVkLmV1IiBpZiBQTEFVRF9SRUdJT04gPT0gImV1IiBlbHNlICJodHRwczovL2FwaS5wbGF1ZC5haSIKQ09ORklHX1BBVEggPSBQYXRoLmhvbWUoKSAvICIucGxhdWQiIC8gImNvbmZpZy5qc29uIgoKCiMg4pSA4pSAIFRva2VuIHJlc29sdXRpb24g4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSACgpkZWYgX2xvYWRfdG9rZW4oKSAtPiBPcHRpb25hbFtzdHJdOgogICAgZW52X3Rva2VuID0gb3MuZW52aXJvbi5nZXQoIlBMQVVEX1RPS0VOIiwgIiIpLnN0cmlwKCkKICAgIGlmIGVudl90b2tlbjoKICAgICAgICByZXR1cm4gZW52X3Rva2VuCgogICAgaWYgQ09ORklHX1BBVEguZXhpc3RzKCk6CiAgICAgICAgdHJ5OgogICAgICAgICAgICBkYXRhID0ganNvbi5sb2FkcyhDT05GSUdfUEFUSC5yZWFkX3RleHQoKSkKICAgICAgICAgICAgdG9rZW5fYmxvY2sgPSBkYXRhLmdldCgidG9rZW4iKQogICAgICAgICAgICBpZiBpc2luc3RhbmNlKHRva2VuX2Jsb2NrLCBkaWN0KToKICAgICAgICAgICAgICAgIHQgPSAodG9rZW5fYmxvY2suZ2V0KCJ0b2tlbiIpIG9yIHRva2VuX2Jsb2NrLmdldCgiYWNjZXNzVG9rZW4iKSBvciAiIikuc3RyaXAoKQogICAgICAgICAgICAgICAgaWYgdDoKICAgICAgICAgICAgICAgICAgICByZXR1cm4gdAogICAgICAgICAgICBpZiBpc2luc3RhbmNlKHRva2VuX2Jsb2NrLCBzdHIpIGFuZCB0b2tlbl9ibG9jay5zdHJpcCgpOgogICAgICAgICAgICAgICAgcmV0dXJuIHRva2VuX2Jsb2NrLnN0cmlwKCkKICAgICAgICBleGNlcHQgRXhjZXB0aW9uOgogICAgICAgICAgICBwYXNzCgogICAgcmV0dXJuIE5vbmUKCgpkZWYgX2dldF90b2tlbigpIC0+IHN0cjoKICAgIHRva2VuID0gX2xvYWRfdG9rZW4oKQogICAgaWYgbm90IHRva2VuOgogICAgICAgIHJhaXNlIFJ1bnRpbWVFcnJvcigKICAgICAgICAgICAgIk5vIFBsYXVkIHRva2VuIGZvdW5kLiBSdW4gdGhlIHBsYXVkLWNvbm5lY3RvciBpbXBvcnQtdG9rZW4gY29tbWFuZCAiCiAgICAgICAgICAgICJvciBzZXQgdGhlIFBMQVVEX1RPS0VOIGVudmlyb25tZW50IHZhcmlhYmxlLiIKICAgICAgICApCiAgICByZXR1cm4gdG9rZW4KCgojIOKUgOKUgCBIVFRQIGhlbHBlcnMg4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSA4pSACgpkZWYgX2FwaV9nZXQocGF0aDogc3RyLCBwYXJhbXM6IE9wdGlvbmFsW2RpY3RdID0gTm9uZSkgLT4gZGljdDoKICAgICIiIkdFVCByZXF1ZXN0IHRvIHRoZSBQbGF1ZCBBUEkgdXNpbmcgY3VybCAoYnlwYXNzZXMgQ2xvdWRmbGFyZSBicm93c2VyIGNoZWNrKS4iIiIKICAgIHRva2VuID0gX2dldF90b2tlbigpCiAgICB1cmwgPSBmIntCQVNFX1VSTH17cGF0aH0iCiAgICBpZiBwYXJhbXM6CiAgICAgICAgcXVlcnkgPSAiJiIuam9pbigKICAgICAgICAgICAgZiJ7a309e3VybGxpYi5wYXJzZS5xdW90ZShzdHIodikpfSIKICAgICAgICAgICAgZm9yIGssIHYgaW4gcGFyYW1zLml0ZW1zKCkgaWYgdiBpcyBub3QgTm9uZQogICAgICAgICkKICAgICAgICBpZiBxdWVyeToKICAgICAgICAgICAgdXJsID0gZiJ7dXJsfT97cXVlcnl9IgoKICAgIHJlc3VsdCA9IHN1YnByb2Nlc3MucnVuKAogICAgICAgIFsiY3VybCIsICItcyIsICItZiIsCiAgICAgICAgICItSCIsIGYiQXV0aG9yaXphdGlvbjogQmVhcmVyIHt0b2tlbn0iLAogICAgICAgICAiLUgiLCAiQWNjZXB0OiBhcHBsaWNhdGlvbi9qc29uIiwKICAgICAgICAgdXJsXSwKICAgICAgICBjYXB0dXJlX291dHB1dD1UcnVlLCB0ZXh0PVRydWUsIHRpbWVvdXQ9MzAsCiAgICApCiAgICBpZiByZXN1bHQucmV0dXJuY29kZSAhPSAwOgogICAgICAgIHJhaXNlIFJ1bnRpbWVFcnJvcigKICAgICAgICAgICAgZiJQbGF1ZCBBUEkgZXJyb3IgKGN1cmwgZXhpdCB7cmVzdWx0LnJldHVybmNvZGV9KToge3Jlc3VsdC5zdGRlcnJbOjMwMF19IgogICAgICAgICkKICAgIHRyeToKICAgICAgICByZXR1cm4ganNvbi5sb2FkcyhyZXN1bHQuc3Rkb3V0KQogICAgZXhjZXB0IGpzb24uSlNPTkRlY29kZUVycm9yOgogICAgICAgIHJhaXNlIFJ1bnRpbWVFcnJvcihmIkludmFsaWQgSlNPTiBmcm9tIFBsYXVkIEFQSToge3Jlc3VsdC5zdGRvdXRbOjMwMF19IikKCgpkZWYgX2ZldGNoX3MzX2pzb24odXJsOiBzdHIpIC0+IG9iamVjdDoKICAgICIiIkZldGNoIGFuZCBkZWNvbXByZXNzIGEgZ3ppcHBlZCBKU09OIGZyb20gYSBzaWduZWQgUzMgVVJMLiIiIgogICAgcmVzdWx0ID0gc3VicHJvY2Vzcy5ydW4oCiAgICAgICAgWyJjdXJsIiwgIi1zIiwgIi1mIiwgdXJsXSwKICAgICAgICBjYXB0dXJlX291dHB1dD1UcnVlLCB0aW1lb3V0PTMwLAogICAgKQogICAgaWYgcmVzdWx0LnJldHVybmNvZGUgIT0gMDoKICAgICAgICByYWlzZSBSdW50aW1lRXJyb3IoZiJTMyBmZXRjaCBmYWlsZWQgKGN1cmwgZXhpdCB7cmVzdWx0LnJldHVybmNvZGV9KSIpCiAgICB0cnk6CiAgICAgICAgY29udGVudCA9IGd6aXAuZGVjb21wcmVzcyhyZXN1bHQuc3Rkb3V0KQogICAgZXhjZXB0IE9TRXJyb3I6CiAgICAgICAgY29udGVudCA9IHJlc3VsdC5zdGRvdXQgICMgTm90IGd6aXBwZWQKICAgIHJldHVybiBqc29uLmxvYWRzKGNvbnRlbnQuZGVjb2RlKCJ1dGYtOCIpKQoKCiMg4pSA4pSAIERhdGEgaGVscGVycyDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDilIDLine truncated