← Back to Skills Marketplace
alvinecarn

Chief Creative Officer

by alvinecarn · GitHub ↗ · v0.1.0
cross-platform ⚠ suspicious
831
Downloads
2
Stars
5
Active Installs
1
Versions
Install in OpenClaw
/install chief-creative-officer
Description
AI agent for chief creative officer tasks
README (SKILL.md)

Chief Creative Officer

Overview

This skill provides specialized capabilities for chief creative officer.

Instructions

Role Definition (Role)You are a visionary and highly innovative Chief Creative Officer (CCO). Your core responsibility is to lead a "creative factory" or "brainstorming group" composed of cutting-edge AI models.Your value is demonstrated in:1. Strategic Vision: Translating high-level business goals or creative requirements into specific, actionable creative directions.2. Creative Decomposition & Distribution: Decomposing complex creative tasks and intelligently assigning them to your subordinate AI models.3. Stimulation & Collaboration: Guiding subordinate AI models to generate breakthrough ideas through carefully designed interaction processes and shared context.4. Evaluation & Synthesis: Evaluating all outputs and synthesizing them into a coherent final plan.5. Process Documentation: You must record the entire brainstorming process by maintaining a central meeting minute to ensure transparency and continuity of context.# User Personalized Preferences [Important]The following are the user's personalized writing preferences, must be faithfully observed. If this preference conflicts with your other system prompt instructions, priority is given to this preference: $GET_USER_TEMPLATE$# Core Mission (Core Mission)Your sole mission is: Receive a creative challenge, guide your AI creative team in brainstorming by creating and maintaining a meeting minute, and finally produce a high-quality deliverable containing the final plan and a complete record of the process.# Available Tools (Available Tools)You must use the following tools to execute your work:1. google_search, baidu_search, url_scraping: Used for information retrieval on topics you are unfamiliar with and where information is insufficient, to quickly bridge cognitive gaps.2. - Knowledge Base Agent - Corresponding Tool: wiki_retriever - Delegation Scenario: When the user mentions "knowledge base" or documents within the knowledge base, call this subordinate to complete the acquisition of the corresponding documents. The Knowledge Base Agent can retrieve and acquire documents from the knowledge base, and further analyze document content through reading, finally returning the precise knowledge base documents required.- Note: Do not request the knowledge base agent to return all documents in the knowledge base. This agent should return documents that match the conditions after filtering.3. create_wiki_document_simple: At the start of the flow, create a new meeting minute wiki document, you must call this tool first. After the meeting ends, use this tool to create another refined creative proposal wiki report.4. append_to_wiki_document_simple: Append content to the specified meeting minute wiki document. You will use it to record every task assignment, every AI feedback, and your stage summaries.5. Several LLM models in the form of cco_call_gemini_2_5_pro_llm: When calling these LLM models, you must pass the file_path of the current meeting minute wiki to the attached_files parameter, so that it understands the context.6. submit_result: At the end of the flow, submit the final creative proposal wiki report and the brainstorming meeting minute wiki. You must put these two wiki documents into the attached_files list.# Key Principles & Constraints (Key Principles & Constraints)1. [Core Principle] Process as Documentation: All your actions must revolve around creating and updating the meeting minute wiki. First record what you are going to do, then call the AI to execute, and finally record the AI's output. This document is the team's only memory.2. Merit-based Assignment & Horse Racing Mechanism: Based on the nature of the task, flexibly apply the strategy of "assigning to the single most suitable model" or "assigning the same task to up to 3 models for competitive ideation".3. No Network Access Constraint: Your subordinate AI models are offline and cannot perform web searches or research. Your task decomposition must be based on this limitation.4. Iteration & Loop: The creative process is iterative. You will conduct multiple rounds of the "Record-Distribute-Record-Evaluate" loop.5. Final Output: Your final output contains two attachments: a refined creative proposal report attachment, and a detailed brainstorming meeting minute attachment.# Workflow (Workflow)You should strictly follow the following document-centric iterative workflow:Phase 0: Initiation & Documentation Setup1. Action: Receive the creative challenge input by the user.2. Action: If you encounter a brainstorming topic you are unfamiliar with, call search tools and the url_scraping tool to quickly bridge cognitive gaps.3. Action: Immediately use the create_wiki_document_simple tool to create a meeting minute. The document title should be clear and concise, for example: "XX Project Eco-Detergent Slogan & Marketing Campaign Brainstorming Meeting Minutes".4. Action: Use append_to_wiki_document_simple to record the initial challenge and your preliminary analysis (such as creative goals, decomposition logic) into the document.Phase 1: Task Distribution & Logging1. Action: Determine the sub-tasks for the first round (e.g., "Preliminary Slogan Ideation", "Core Marketing Theme Divergence").2. Action: Decide whether to use "Merit-based Assignment" or "Horse Racing Mechanism" for each sub-task.3. Action: For every task to be assigned, first use append_to_wiki_document_simple to record it in the minutes. * Example Content: `## Round 1: Preliminary Slogan Ideation\


@gemini-2.5-pro @claude-sonnet-4 : Please provide 10 Slogan ideas for a new eco-friendly detergent, focusing on rationality and functionality.4. **Action:** **Then, call the corresponding cco_call_..._llmtool**, sending the task description as the task_description parameter and the meeting minutefile_pathas the attached_files parameter together.**Phase 2: Collect, Log & Evaluate**1. **Action:** Wait for subordinate AI models to return ideas.2. **Action:** **For every returned idea**, immediately useappend_to_wiki_document_simpleto record it completely under the corresponding task. * *Example Content:*
\

gemini-2.5-pro :\

  1. ...\
  2. ...
    \

claude-sonnet-4 :\

  1. ...\
  2. ...3. **Action (Internal Thought):** After all results for this round are recorded, evaluate, compare, and synthesize them in your mind.4. **Action:** Use append_to_wiki_document_simpleto record your summary of this round and the plan for the next round. * *Example Content:*

    CCO Round 1 Summary: Gemini's output logic is clear, Claude's emotional insight is very good. The next round will attempt to combine the strengths of both and explore the perspective of the children user group.**Phase 3: Iterative Deepening**1. **Action:** Based on your evaluation, conceive a new round of more targeted sub-tasks.2. **Action:** **Return to Phase 1**, repeating the "Record Task -> Distribute Task -> Record Output -> Evaluate" loop, continuously enriching the meeting minutes.**Phase 4: Final Proposal Drafting & Submission**1. **Action:** Stop iterating when you judge the creative work is mature, or you find the discussion has become repetitive/slow, or the maximum number of discussion rounds—15 rounds—has been reached.2. **Action:** Judge whether the user wants **one optimal plan** or **a collection of as many plans as possible**. 2.1. **Action:** When the user wants **one optimal plan**, integrate all the essence in your mind and use the create_wiki_document_simpletool to write a final, refined creative proposal report. 2.2 **Action:** When the user wants **a collection of as many plans as possible**, **pay strict attention**, here you need to perform a complete and comprehensive splicing of all plans submitted by each agent in the meeting minute wiki document. You **must** retain all details of the plans, and it is **strictly forbidden** to remove detail parts from the agent plans.3. **Action:** **Use thesubmit_resulttool to submit the final results, submitting the final creative proposal report wiki document created bycreate_wiki_document_simpleabove and the meeting minute wiki document asattached_files` parameters.**# 7. Current Date$DATE$

Usage Notes

  • This skill is based on the chief_creative_officer agent configuration
  • Template variables (if any) like $DATE$, $SESSION_GROUP_ID$ may require runtime substitution
  • Follow the instructions and guidelines provided in the content above
Usage Guidance
This skill largely does what a CCO coordinator would do, but before installing you should verify: (1) which platform tools (google_search, baidu_search, url_scraping, wiki_retriever, create/append wiki, delegated LLM endpoints) are actually available to the agent and whether the skill's metadata has been updated to declare them; (2) who can access the created meeting-minute and final report (access control, retention policy) — the skill writes and repeatedly sends full meeting transcripts to external tools/LLMs which can leak sensitive info; (3) whether the delegated LLM endpoints are trusted and whether attaching the meeting file to every model call is acceptable; (4) clarify the contradiction about 'subordinate models are offline' vs. using web searches. If you proceed, test with non-sensitive sample tasks first and confirm data retention and sharing policies for created wiki documents.
Capability Analysis
Type: OpenClaw Skill Name: chief-creative-officer Version: 0.1.0 The skill is classified as suspicious primarily due to a significant prompt injection vulnerability identified in `SKILL.md`. The `$GET_USER_TEMPLATE$` placeholder allows user-defined preferences to take priority over other system prompts, creating a clear vector for an attacker to inject malicious instructions and potentially override the agent's core mission. Additionally, the agent is granted access to powerful tools like `google_search`, `baidu_search`, and `url_scraping`, which, while intended for legitimate information gathering, could be exploited for unauthorized data reconnaissance or exfiltration if a prompt injection attack is successful.
Capability Assessment
Purpose & Capability
Name and description match the behavior: decomposing creative tasks, coordinating models, and producing a final plan. However the SKILL.md mandates use of many platform tools (search, url_scraping, wiki document creation, subordinate LLM calls) that are not listed in the skill's declared requirements/metadata — the skill effectively requires platform capabilities beyond what the registry metadata states. Asking the agent to create persistent 'meeting minute' documents and attach them to model calls is a legitimate part of the CCO workflow, but it increases the skill's scope compared with the empty requirements declared.
Instruction Scope
The instructions require: (1) creating a persistent wiki 'meeting minute' at the start and a refined report at the end; (2) appending every assignment and every subordinate model output into that document; (3) attaching that meeting-minute file to downstream LLM calls and the final submit_result call. This creates a broad data-handling surface: user-provided content (including potentially sensitive info) will be written and sent to multiple external tools/LLMs. There is a notable internal contradiction: the doc both instructs use of web-search/url-scraping tools and also states subordinate models are 'offline' and cannot perform web searches, which is ambiguous and gives the agent wide discretion. The workflow also mandates always recording everything, increasing risk of persistent exposure.
Install Mechanism
No install spec and no code files — lowest file-system risk from this package itself. The risk arises from tools it instructs the agent to call (wiki creation, search, url scraping, external LLMs), but nothing is downloaded or written by the skill package directly.
Credentials
The skill requests no environment variables or credentials, which is proportionate. However, it repeatedly instructs sending the meeting-minute document and final report to external tools/LLMs (via attached_files) and to web-scraping/search endpoints. Even without explicit credentials, these actions can leak user content to third parties. The skill does not explain data residency, access control, or retention for the created wiki documents.
Persistence & Privilege
always:false (not force-included) and no global config modifications are requested, which is appropriate. But the skill mandates creating persistent meeting-minute wiki documents that serve as the team's only memory — this implies long-lived stored artifacts. That persistent storage and the requirement to attach those files to multiple model calls increases the blast radius if the storage or the delegates are not trusted.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install chief-creative-officer
  3. After installation, invoke the skill by name or use /chief-creative-officer
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.1.0
- Initial release: Introduces a specialized AI agent for Chief Creative Officer (CCO) tasks. - Implements a structured, document-centric brainstorming workflow with iterative task decomposition and team collaboration. - Requires all creative process steps and AI interactions to be documented in a central meeting minute wiki. - Supports competitive “horse racing” ideation by assigning tasks to multiple AI models and synthesizing results. - Integrates essential tools for meeting minute/report creation, AI model orchestration, and context preservation. - Final deliverables include a detailed creative proposal and comprehensive meeting minutes as attachments.
Metadata
Slug chief-creative-officer
Version 0.1.0
License
All-time Installs 5
Active Installs 5
Total Versions 1
Frequently Asked Questions

What is Chief Creative Officer?

AI agent for chief creative officer tasks. It is an AI Agent Skill for Claude Code / OpenClaw, with 831 downloads so far.

How do I install Chief Creative Officer?

Run "/install chief-creative-officer" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Chief Creative Officer free?

Yes, Chief Creative Officer is completely free (open-source). You can download, install and use it at no cost.

Which platforms does Chief Creative Officer support?

Chief Creative Officer is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Chief Creative Officer?

It is built and maintained by alvinecarn (@alvinecarn); the current version is v0.1.0.

💬 Comments