Home / All Skills / Content & Media Creation / slack-gif-creator
English (EN)简体中文 (ZH)日本語 (JA)
Content & Media Creation · anthropics/skills

slack-gif-creator

Knowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack."

L4 High-Privilege · Graded auth & full auditV1 Commit-PinnedExecute CommandWrite FileDual-pass Audit
Author anthropicsGitHub anthropics/skills ↗License Apache-2.0（skill 内 LICENSE.txt：Copyright 2026 Anthropic, PBC.）Commit 41bbe19d1a
Control levels describe skill capabilities and suggested control measures; they do not represent quality or safety certifications, nor do they dictate usability. Usage decisions rest entirely with users and administrators. Different agent hosts apply varying constraints; do not assume hosts provide automatic protection. Third-party skill copyrights belong to their original authors. Use at your own risk.

1How It Works & Mechanism

SKILL.md focuses primarily on "Knowledge + Orchestration": it hardcodes Slack specification constraints into the prompt, guides the agent to hand-draw animations frame-by-frame using PIL primitives, and then delegates them to bundled Python tools for encoding and optimization.

SKILL.md
# Slack GIF Creator A toolkit providing utilities and knowledge for creating animated GIFs optimized for Slack. ## Slack Requirements **Dimensions:** - Emoji GIFs: 128x128 (recommended) - Message GIFs: 480x480
Dimensions, FPS, color count, and duration are prepended as hard constraints at the head of SKILL.md and injected into the agent's context; all drawing relies on the agent writing code with ImageDraw primitives (no built-in assets or templates).

Drawing capabilities are entirely delegated to the running Python + Pillow environment: SKILL.md explicitly requires the agent to write PIL drawing code on the fly, and the skill provides no pre-made graphics.

SKILL.md
## Philosophy This skill provides: - **Knowledge**: Slack's requirements and animation concepts - **Utilities**: GIFBuilder, validators, easing functions - **Flexibility**: Create the animation logic using PIL primitives It does NOT provide: - Rigid animation templates or pre-made functions - Emoji font rendering (unreliable across platforms) - A library of pre-packaged graphics built into the skill
Core design: Graphic creativity is generated on the fly by the model through code (high degree of freedom, but output quality per generation depends on model performance); the skill only guarantees output-side encoding and validation quality.

GIFBuilder is the true core engine: it takes PIL/numpy frame sequences, performs palette quantization, deduplication, automatic emoji-mode scaling/frame dropping, and finally writes to disk using imageio.

core/gif_builder.py
def save( self, output_path: str | Path, num_colors: int = 128, optimize_for_emoji: bool = False, remove_duplicates: bool = False, ) -> dict: ... # Optimize colors with global palette optimized_frames = self.optimize_colors(num_colors, use_global_palette=True) # Calculate frame duration in milliseconds frame_duration = 1000 / self.fps # Save GIF imageio.imwrite( output_path, optimized_frames, duration=frame_duration, loop=0, # Infinite loop )
optimize_colors uses sampled frames stitched together to generate a global palette (method=2) and applies frame-by-frame quantize; emoji mode downscales frames larger than 128x128 to 128x128, caps color count at 48, and thins frame count down to ~12.

validators reads back generated GIFs using PIL for independent quality checks, forming a closed "generation → self-check" loop.

core/validators.py
def validate_gif( gif_path: str | Path, is_emoji: bool = True, verbose: bool = True ) -> tuple[bool, dict]: """ Validate GIF for Slack (dimensions, size, frame count). ... with Image.open(gif_path) as img: width, height = img.size # Count frames frame_count = 0 try: while True: img.seek(frame_count) frame_count += 1 except EOFError: pass
Counts frames by seeking frame-by-frame and derives fps using info['duration']; note that the actual passing thresholds do not strictly align with the hard specifications claimed in the SKILL.md header (see verification table).

Animation concepts (shake/pulse/bounce/spin/fade/slide/zoom/explode) are injected purely via SKILL.md text descriptions, implemented on the fly by the agent using easing utility functions + math.

SKILL.md
### Bounce Object falls and bounces: - Use `interpolate()` with `easing='bounce_out'` for landing - Use `easing='ease_in'` for falling (accelerating) - Apply gravity by increasing y velocity each frame
easing.py provides the full suite of linear/quad/cubic/bounce/elastic/back alongside two additional utilities, squash_stretch and arc_motion, written in pure math without dependencies.

2Core Capabilities

01Frame sequence assembly and Slack-targeted GIF encoding (global palette quantization, infinite looping)
02One-click emoji optimization: automatically scales down frames >128x128 to 128x128, caps colors at 48, and thins frames to ~12
03Near-duplicate frame deduplication (similarity threshold 0.9995 to preserve subtle animations)
04GIF self-inspection: readback validation of dimensions, file size, frame count, duration, and fps (separate thresholds for emoji and message)
0517+ easing functions and interpolations (linear/quad/cubic/bounce/elastic/back, including aliases anticipate/overshoot)
06Frame drawing helpers: blank frames, vertical gradient backgrounds, circles, stars, and text
07User-uploaded image handling guidelines (direct use, using as inspiration, frame cropping)

3External Dependencies

KindDependency / Endpoint
packagepillow (PIL)
packageimageio（v3 API，GIF 写盘）
packageimageio-ffmpeg
packagenumpy（帧为 np.ndarray、相似度/像素堆叠计算）

4Recommended Governance & Evidence

Suggested Controls for Safe Operation
For Agent Runtime
Graded/per-use authorization, sandbox recommended
For Administrators
Approval workflow, rollback support, full audit
Evidence & Triggers (File Paths & Excerpts)
  • scripts/: Execute local scripts or commands: The host-side skill usage is "agent generates Python code on the fly based on SKILL.md and executes it" (SKILL.md is full of API examples, no __m
  • 本地工作目录: File read/write and artifact output: GIFBuilder.save() writes GIF files according to the output_path parameter (path is provided by agent-generated code and can be any path)
Audit Notes: Pure local code generation + image library toolkit: no network calls, no credential access, no hidden/obfuscated scripts, no dynamic execution of external input, no template injection. File writes occur solely through GIFBuilder.save (which is the core function of the skill, with paths controlled by the caller). The only considerations: the host must allow the agent to execute Python; parsing user-uploaded images depends on the security patch status of Pillow (dependency-level risk, not a skill defect).

5Second-Pass Verification

  • [ok] 外部依赖 pillow — requirements 声明 pillow>=10.0.0；gif_builder.py/frame_composer.py/validators.py 均真实 `import`/`from PIL import`，调用点充分
  • [ok] 外部依赖 imageio（GIF 编码） — requirements 声明 imageio>=2.31.0；gif_builder.py:12 `import imageio.v3 as imageio`，save() 中 imageio.imwrite 真实调用
  • [discrepancy] 外部依赖 imageio-ffmpeg — requirements.txt:3 声明 imageio-ffmpeg>=0.4.9，但 4 个核心 py 文件均无直接 import；imageio 写 GIF 走自有/ Pillow 后端，不必然需要 ffmpeg。属冗余或为 imageio 其他格式兜底，仅作防御性声明。影响：安装清单略夸大
  • [ok] 外部依赖 numpy — gif_builder.py/frame_composer.py 真实 import numpy；帧去重(np.abs/mean)、调色板采样(np.vstack)确有调用
  • [ok] 无网络调用/凭据读取/隐藏执行 — 第二遍全目录正则复查零命中；无隐藏文件、无子模块、无 __main__/argparse；easing.py 仅 import math
  • [discrepancy] 『Slack Requirements』与 validators 门槛一致性 — SKILL.md 头部写 Emoji 128x128 / Message 480x480、FPS 10-30、Colors 48-128；validators.py 实际门槛为 emoji 64-128 方形通过、message 仅查宽高比≤2 且短边 320-640——与『480x480』宣称不符；且均未给出 Slack 官方文档出处，属启发式规格。另 builder.save() 默认 num_colors=128 超出文档推荐的 48-128 上限（128 是上限本身，相符）；emoji 优化强制 48 色却把帧数砍到 ~12（fps 名义不变，时长被压缩），与头部『FPS 10-30』经验值存在偏差
  • [ok] SKILL.md 功能声明 vs 代码能力相符性 — 声明的 GIFBuilder/validators/easing/frame_composer 四模块全部真实存在且 API 签名与文档一致；create_gradient_background、draw_star 等均实现；draw_text 用 load_default() 位图字体与『不用 emoji 字体』哲学一致；『remove_duplicates=True』去重阈值 0.9995 与 docstring 相符
  • [ok] 帧/尺寸写入行为 — add_frame 会把非目标尺寸帧 LANCZOS 重采样到 builder 尺寸；save 返回 size_kb/size_mb/dimensions/frame_count/fps/duration_seconds/colors 信息字典并打印 ✓ 摘要——与文档一致

6Verdict & Scenarios

  • Clean mechanics, single-purpose: Knowledge (prompt) + 4 small libraries, no template magic, graphical creativity is completely open for the model to unleash
  • Encoding pipeline genuinely optimized for compression: global palette (cross-frame sampling collage before quantize) rather than per-frame local paletting, yielding superior compression
  • One-click compliance for emoji scenarios: automatically resizes to 128x128, caps colors at 48, and drops frames to ~12, operationalizing Slack emoji size constraints
  • Generation + self-check closed loop: validate_gif reads back finished artifacts, seeking frame-by-frame to count frames and derive fps, preventing broken outputs upon generation
  • Comprehensive animation math: 17+ easings + squash_stretch + parabolic arc_motion, covering natural motions like bounce/elastic/back without requiring the agent to derive formulas manually
    • Best for: Best for: Scenarios where users want Claude-family agents to produce "upload-ready emoji/message GIFs" in Slack (team internal emoji packs, release alert animations, rapid prototyping); most compatible with code-generating hosts (Claude Code/Codex types); requires local Python3 + Pillow/imageio/numpy. Also serves human developers as a quick PIL animation cheat sheet + optimization pipeline library.
    Not for: Not suitable for: (1) Hosts without code execution capability or Python environments (limited value at the pure knowledge level); (2) Requirements needing photorealistic/complex 3D rendering or calling real image/video APIs (e.g., DALL·E frame generation, ffmpeg video-to-GIF)—this skill only draws vector primitives and does not touch external generation services; (3) Users requiring officially certified compliant delivery—specifications are heuristic and must be manually verified against current Slack constraints.
    Install Agent-Ready
    ① Verified Mirror 2026-09-06
    Option A · Direct DownloadDownload slack-gif-creator.tar.gz
    sha256: 9d54ca459d0b781f…
    Option B · Copy Guide for Agent
    install.json
    Agents automatically download and verify mirrors from this JSON guide.
    ② Upstream GitHub
    GitHub Repo ↗
    Pinned to commit 41bbe19d1a.