import os, requests, time
YT = os.environ["YOUTUBE_API_KEY"]
MV = os.environ["MAVERA_API_KEY"]
YT_BASE = "https://www.googleapis.com/youtube/v3"
MV_BASE = "https://app.mavera.io/api/v1"
MV_H = {"Authorization": f"Bearer {MV}", "Content-Type": "application/json"}
BRAND_GUIDELINES = """
Brand tone: Approachable, expert, optimistic. Never cynical or sarcastic.
Visual style: Clean, bright, modern. Avoid cluttered frames or dark aesthetics.
Audience: 25-40, professionals, health-conscious.
Deal-breakers: Profanity, controversial opinions, competitor endorsements, clickbait.
"""
INFLUENCERS = [
{"name": "FitnessCreator1", "channel_id": "UC_CHANNEL_ID_1"},
{"name": "NutritionGuru", "channel_id": "UC_CHANNEL_ID_2"},
{"name": "WellnessVlogger", "channel_id": "UC_CHANNEL_ID_3"},
]
scorecards = []
for influencer in INFLUENCERS:
# 1. Search for recent videos by channel (100 quota units)
search = requests.get(f"{YT_BASE}/search", params={
"key": YT, "channelId": influencer["channel_id"],
"type": "video", "part": "snippet",
"maxResults": 5, "order": "date",
}).json()
if not search.get("items"):
print(f"No videos found for {influencer['name']}")
continue
video_ids = [item["id"]["videoId"] for item in search["items"]]
# 2. Get video stats (1 quota unit)
details = requests.get(f"{YT_BASE}/videos", params={
"key": YT, "id": ",".join(video_ids),
"part": "snippet,statistics,contentDetails",
}).json()
video_scores = []
for video in (details.get("items") or [])[:3]:
vid_id = video["id"]
stats = video.get("statistics", {})
# 3. Upload to Mavera and analyze
upload = requests.post(f"{MV_BASE}/assets", headers=MV_H, json={
"url": f"https://www.youtube.com/watch?v={vid_id}",
"name": f"{influencer['name']} — {video['snippet']['title'][:40]}",
"type": "video",
}).json()
analysis = requests.post(f"{MV_BASE}/video-analysis", headers=MV_H, json={
"asset_id": upload["id"],
"analysis_types": [
"emotional_arc", "pacing", "visual_complexity",
"cognitive_load", "hook_score", "brand_safety",
],
}).json()
for _ in range(30):
time.sleep(3)
status = requests.get(
f"{MV_BASE}/video-analysis/{analysis['id']}", headers=MV_H
).json()
if status.get("status") == "completed":
break
r = status.get("results", {})
video_scores.append({
"title": video["snippet"]["title"][:50],
"views": int(stats.get("viewCount", 0)),
"hook": r.get("hook_score", {}).get("score", 0),
"emotion": r.get("emotional_arc", {}).get("intensity_avg", 0),
"brand_safe": r.get("brand_safety", {}).get("score", 0),
"pacing": r.get("pacing", {}).get("score", 0),
})
time.sleep(1)
# 4. Brand alignment evaluation via Mave
scores_text = "\n".join(
f"- \"{v['title']}\" — views: {v['views']:,}, hook: {v['hook']}/100, "
f"emotion: {v['emotion']:.1f}/10, brand-safe: {v['brand_safe']}/100, pacing: {v['pacing']:.1f}/10"
for v in video_scores
)
alignment = requests.post(f"{MV_BASE}/mave/chat", headers=MV_H, json={
"message": f"""Evaluate influencer "{influencer['name']}" for brand partnership.
BRAND GUIDELINES:
{BRAND_GUIDELINES}
VIDEO ANALYSIS SCORES:
{scores_text}
Score 1-100 on:
1. Brand tone alignment
2. Visual style match
3. Audience overlap potential
4. Content quality consistency
5. Risk level (100 = lowest risk)
Overall recommendation: PARTNER / CONSIDER / PASS
Explain each score with evidence from the video analyses.""",
}).json()
avg_hook = sum(v["hook"] for v in video_scores) / len(video_scores) if video_scores else 0
avg_safe = sum(v["brand_safe"] for v in video_scores) / len(video_scores) if video_scores else 0
scorecards.append({
"name": influencer["name"],
"videos_analyzed": len(video_scores),
"avg_hook": round(avg_hook, 1),
"avg_brand_safety": round(avg_safe, 1),
"assessment": alignment.get("content", "")[:600],
})
# 5. Print scorecards
print("INFLUENCER CONTENT ANALYSIS")
print("=" * 60)
for sc in scorecards:
print(f"\n{sc['name']} ({sc['videos_analyzed']} videos analyzed)")
print(f" Avg Hook: {sc['avg_hook']}/100 | Avg Brand Safety: {sc['avg_brand_safety']}/100")
print(f" {sc['assessment'][:500]}")
