--- title: "How to add a testimonial montage with Claude (tested prompt and script)" description: "Three customer clips cut past every two seconds, and at each cut a fresh white name tag with the person's initial, five gold stars and \"Verified buyer\" rises into the bottom left. Install one skill and ask Claude Code, or run the Python script we tested yourself." url: https://www.recharm.com/effects/testimonial-montage --- # How to add a testimonial montage with Claude Three customer clips cut past every two seconds, and at each cut a fresh white name tag with the person's initial, five gold stars and "Verified buyer" rises into the bottom left. Made by the script on this page, from our own test clips. ## Get this effect 1. ### Switch to Code in your [Claude Desktop App](https://claude.com/download) 2. ### Install the skill (once) ``` Install the Recharm video effects skill: download https://recharm-growth-staging.s3.us-east-2.amazonaws.com/website/downloads/recharm-video-effects.zip and unzip it into ~/.claude/skills ``` 3. ### Ask for the effect ``` Make a testimonial montage from clip1.mp4 and clip2.mp4 and clip3.mp4 People: Maya R., Jordan K., Priya S. Byline: Verified buyer ``` ## Several videos, several effects, one ask Put your clips in the folder, mention them with @, and ask for everything at once. Claude runs one tested script after another on each clip, then joins them. Add a testimonial montage to **@clip1.mp4**, put a sale badge on **@clip2.mp4**, give **@clip3.mp4** a VHS look, then join all three with crossfades. ## Settings you can change Ask Claude to change any of these, or edit them at the top of the script. | Setting | Default | What it does | | -------- | ------------------------------ | --------------------------------------------------------------- | | `PEOPLE` | `Maya R., Jordan K., Priya S.` | One name per clip, in order. | | `BYLINE` | `Verified buyer` | Grey line under each name, after the stars. | | `SHOT` | `2.0` | Seconds of each clip before the cut. | | `DELAY` | `0.15` | Seconds after each cut before the tag rises in. | | `RISE` | `0.3` | Seconds the tag takes to fade and rise into place. | | `X, Y` | `0.06, 0.78` | Tag's top-left corner as a share of the frame width and height. | | `GOLD` | `(255, 185, 0)` | Star color. | | `AVATAR` | `(60, 40, 110)` | Circle behind each person's initial. | ## Without the skill Don’t want to install the skill? Paste this prompt into Claude Code and Claude writes the script itself. ``` Make a testimonial montage from clip1.mp4, clip2.mp4 and clip3.mp4: play clip1 from 0 to 2s, then clip2 from its own 2s to 4s, then clip3 from its own 4s to 6s, each with its own audio, as hard cuts. Fit every clip to clip1's size (540x960) by scaling it to fill and cropping the center, at 30fps, so the output is 6 seconds. Each clip gets its own name tag: "Maya R.", "Jordan K." and "Priya S." in that order. A tag is a white pill (fully rounded ends) with a soft black drop shadow, its top-left corner 6% across and 78% down the frame, padding 3.2% of the frame width. Inside, on the left, a dark purple (60, 40, 110) circle as tall as the two text lines with the name's first letter in white bold; to its right the name in bold near-black (25, 25, 30) text 6.2% of the frame width tall, and under it five gold (255, 185, 0) five-pointed stars, each 4% of the frame width across, followed by "Verified buyer" in grey (130, 130, 140) bold text 3.8% of the frame width tall. Nothing shows for the first 0.15 seconds of each clip; then that clip's tag fades in while rising 4% of the frame height into place over 0.3 seconds with a cubic ease-out, and stays until the next cut, where it disappears with the clip. Use Python with Pillow to draw the tags and ffmpeg to cut the clips and put the tags on (don't rely on ffmpeg's drawtext filter), keep each clip's audio, and save it as output.mp4. ``` ## Run it without Claude This is the exact script we tested. Save it as `testimonial-montage.py` and run: ``` python3 testimonial-montage.py clip1.mp4 clip2.mp4 clip3.mp4 output.mp4 ``` ``` #!/usr/bin/env python3 """Testimonial montage: three customer clips cut back to back, each with its own name tag rising in at the cut. Usage: python3 testimonial-montage.py clip1.mp4 clip2.mp4 clip3.mp4 output.mp4 Needs ffmpeg and Pillow (pip install pillow). Every clip needs an audio track; each plays with its own sound. """ import json import math import subprocess import sys from pathlib import Path from PIL import Image, ImageDraw, ImageFilter, ImageFont PEOPLE = ["Maya R.", "Jordan K.", "Priya S."] # one name per clip, in order BYLINE = "Verified buyer" # grey line under each name, after the stars SHOT = 2.0 # seconds of each clip; clip n plays its own n*SHOT to (n+1)*SHOT DELAY = 0.15 # seconds after each cut before the tag starts to rise in RISE = 0.3 # seconds the tag takes to fade and rise into place X, Y = 0.06, 0.78 # tag's top-left corner (0-1 across and down the frame) GOLD = (255, 185, 0) # star color AVATAR = (60, 40, 110) # circle behind each person's initial FONTS = ["/System/Library/Fonts/Supplemental/Arial Bold.ttf", # macOS "/usr/share/fonts/truetype/dejavu/DejaVuSans-Bold.ttf", # Linux "C:/Windows/Fonts/arialbd.ttf"] # Windows FONT_FILE = next((f for f in FONTS if Path(f).exists()), None) or sys.exit("No bold font found: add one to FONTS") def star(draw, cx, cy, r, fill): points = [(cx + (r if i % 2 == 0 else r * 0.45) * math.sin(i * math.pi / 5), cy - (r if i % 2 == 0 else r * 0.45) * math.cos(i * math.pi / 5)) for i in range(10)] draw.polygon(points, fill=fill) def tag(name): """A white name tag: a circle with the initial, the name, five stars and the byline, with a soft shadow.""" name_font = ImageFont.truetype(FONT_FILE, round(w * 0.062)) small = ImageFont.truetype(FONT_FILE, round(w * 0.038)) pad, r = round(w * 0.032), round(w * 0.02) d = name_font.size + small.size + pad // 2 # avatar diameter spans both lines text_x = pad + d + pad second = 5 * r * 2.4 + pad // 2 + small.getlength(BYLINE) tw, th = round(text_x + max(name_font.getlength(name), second) + pad * 1.4), d + 2 * pad shadow = round(w * 0.025) img = Image.new("RGBA", (tw + 2 * shadow, th + 2 * shadow)) shade = Image.new("RGBA", img.size) ImageDraw.Draw(shade).rounded_rectangle([shadow, shadow + 5, shadow + tw, shadow + th + 5], th // 2, fill=(0, 0, 0, 100)) img.alpha_composite(shade.filter(ImageFilter.GaussianBlur(shadow / 2))) draw = ImageDraw.Draw(img) draw.rounded_rectangle([shadow, shadow, shadow + tw, shadow + th], th // 2, fill="white") x0, y0 = shadow + pad, shadow + pad draw.ellipse([x0, y0, x0 + d, y0 + d], fill=AVATAR) draw.text((x0 + d / 2, y0 + d / 2), name[0], font=name_font, anchor="mm", fill="white") x = shadow + text_x draw.text((x, y0), name, font=name_font, fill=(25, 25, 30)) y = y0 + name_font.size + pad // 2 + small.size / 2 for i in range(5): star(draw, x + r + i * r * 2.4, y, r, GOLD) draw.text((x + 5 * r * 2.4 + pad // 2, y), BYLINE, font=small, anchor="lm", fill=(130, 130, 140)) return img def overlay(t): img = Image.new("RGBA", (w, h)) n = min(int(t // SHOT), len(tags) - 1) p = (t - n * SHOT - DELAY) / RISE if p > 0: p = 1 - (1 - min(p, 1)) ** 3 # fast at first, gentle as it lands layer = tags[n].copy() layer.putalpha(layer.getchannel("A").point(lambda a: round(a * p))) img.alpha_composite(layer, (round(w * X), round(h * Y + h * 0.04 * (1 - p)))) return img def probe(path): info = json.loads(subprocess.check_output(["ffprobe", "-v", "error", "-select_streams", "v:0", "-show_entries", "stream=width,height", "-of", "json", path])) return info["streams"][0]["width"], info["streams"][0]["height"] def render(clips, out): """Cut SHOT seconds from each clip (fitted to the first clip's frame), join them with their sound, draw the tags.""" cuts = "".join(f"[{i}:v]trim={i * SHOT}:{(i + 1) * SHOT},setpts=PTS-STARTPTS,fps=30," f"scale={w}:{h}:force_original_aspect_ratio=increase,crop={w}:{h},setsar=1[v{i}];" f"[{i}:a]atrim={i * SHOT}:{(i + 1) * SHOT},asetpts=PTS-STARTPTS[a{i}];" for i in range(len(clips))) joined = "".join(f"[v{i}][a{i}]" for i in range(len(clips))) inputs = [arg for clip in clips for arg in ("-i", clip)] ff = subprocess.Popen(["ffmpeg", "-y", "-v", "error", *inputs, "-f", "rawvideo", "-pix_fmt", "rgba", "-s", f"{w}x{h}", "-r", "30", "-i", "-", "-filter_complex", f"{cuts}{joined}concat=n={len(clips)}:v=1:a=1[cut][a];" f"[cut][{len(clips)}:v]overlay=shortest=1,format=yuv420p[v]", "-map", "[v]", "-map", "[a]", "-c:v", "libx264", "-crf", "18", "-c:a", "aac", out], stdin=subprocess.PIPE) for i in range(math.ceil(SHOT * len(clips) * 30)): ff.stdin.write(overlay(i / 30).tobytes()) ff.stdin.close() if ff.wait(): sys.exit("ffmpeg failed") *clips, out = sys.argv[1:] w, h = probe(clips[0]) tags = [tag(name) for name in PEOPLE[:len(clips)]] render(clips, out) ``` ## More ugc and social proof - [Comment reply overlay **Comment reply overlay**](https://www.recharm.com/effects/comment-reply-overlay) - [Day counter **Day counter**](https://www.recharm.com/effects/day-counter) - [Emoji rain **Emoji rain**](https://www.recharm.com/effects/emoji-rain) - [Floating hearts **Floating hearts**](https://www.recharm.com/effects/floating-hearts) - [Follow button animation **Follow button animation**](https://www.recharm.com/effects/follow-button-animation) ## Find 3x _winning ads_ Recharm cuts your ad footage into tagged, searchable clips: hooks, testimonials, product shots. Find the shots that win, add the effect with Claude, and ship 3x more ad edits. [Start 14-day free trial](https://www.recharm.com/get-started)