菜单
Deterministic feedback processor. Fetches client feedback emails from Gmail, downloads attachments (screenshots), and prepares a structured report for triaging into GitHub issues.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Generate, edit, describe, restyle, restore, thumbnail, and batch-process images using xAI (Grok) or OpenAI image APIs and GPT-4o vision. Default provider is xAI ($0.02/image flat rate). Use this skill whenever the user asks to generate, create, make, draw, or design an image or picture using AI, or wants to edit, modify, transform, restyle, composite, or inpaint an existing image. Also handles image description and alt-text generation, background removal, style transfer, photo restoration, thumbnail creation, and batch generation from JSON manifests. Trigger when the user mentions DALL-E, gpt-image, Grok image, xAI image, OpenAI image generation, or wants AI-generated visuals for any purpose (logos, mockups, illustrations, thumbnails, icons, concept art, memes). Also trigger for batch image generation, generating a set or series of images, processing multiple images from a manifest, or creating consistent image collections. If the user says "make me an image of...", "generate a picture", "edit this photo to..
Sets up, configures, and optimizes Google Analytics 4 (GA4) properties. Evaluates websites for proper GA4 implementation, tracking codes, and configuration improvements. Uses the Google Analytics Admin API for programmatic setup or provides manual integration paths via gtag.js or Next.js Third Parties.
Agentic OS Orchestrator. Process and execute tasks from the shared .agent/state/tasks.json queue. Use when the user asks to 'check the queue', 'process tasks', or run the heartbeat.
Framework-agnostic persistent memory and self-improvement loops for AI agents. Scaffolds shared state, task queues, and learnings files that can be read/written by Claude, Gemini, and Antigravity. Use this to initialize an Agentic OS layer in any workspace and instruct agents on how to use it.
Script-First design system workflow. Uses a deterministic auditor script to detect frameworks, inventory components, and detect design drift (hardcoded hex, arbitrary Tailwind). Provides a clear report for the agent to judge and implement a Kitchen Sink page.
Script-First llms.txt generator. Uses a deterministic script to crawl the project structure, identify brand guides, and catalog content files. Provides a repo manifest for the agent to draft context-aware /llms.txt and /llms-full.txt files.
| name | client-feedback |
| description | Deterministic feedback processor. Fetches client feedback emails from Gmail, downloads attachments (screenshots), and prepares a structured report for triaging into GitHub issues. |
| version | 2.0.0 |
Transform client feedback emails into tracked, investigated, and resolved GitHub issues with professional response emails.
This workflow uses a Script-First approach: a deterministic Python script handles the "labor" of fetching and parsing emails, leaving the agent to handle the "judgment" of triaging and fixing issues.
When the user asks to "process feedback", "check emails from [domain]", or "turn feedback into issues":
Run the processor script:
python3 ~/.agents/skills/client-feedback/scripts/process_feedback.py \
--domain "<client-domain>" \
--days 7 \
--out "./feedback-batch-$(date +%F)"
Read the report:
Load the generated report.json and the full body text files to understand the feedback.
Triage to GitHub: For each distinct piece of feedback in the report, create a GitHub issue or update an existing one.
The process_feedback.py script performs these tasks in seconds:
report.json summarizing threads, participants, and snippets.GitHub issues are the primary work board. One issue per item, not one per email.
## Reported by
[Person name] — [date] "[email subject]" (thread [threadId])
## Issue
[Verbatim quote if short, paraphrase if long]
## Status
[Open / awaiting client / in progress]
Apply labels based on category: feedback (always), bug, content, design, provider-data, question.
Verify the feedback against the actual codebase. Grep globally for subjects of claims to ensure full removal or update.
Draft professional responses organized by email thread.
Key Rule: ALWAYS show the draft body text in chat before sending.
Keep signatures simple and professional. Avoid disclosing AI co-authorship unless the client relationship explicitly welcomes it.
If you need to draft a reply with inline images (e.g. before/after proofs), use the gws multipart upload flow documented in the script comments or process_feedback.py logic.