Research Idea

Launch background Clawdbot sessions to explore and analyze business ideas. Say 'Idea: [description]' to trigger. Fork of 'idea' skill rewritten to use sessions_spawn instead of claude CLI + tmux + telegram CLI. Results sent to current chat, not Saved Messages. Zero external dependencies.

This skill appears to do exactly what it says: spawn background Clawdbot sessions, run web searches, write a research.md under ~/clawd/ideas/<slug>/, and send the result back to the chat. Before installing or enabling it, consider: 1) Do you trust the Clawdbot instance and its messaging integration? The skill will use the platform's existing ability to post to Telegram (no new tokens are requested, but the platform's bot credentials will be used). 2) The skill writes files to your home directory (~/clawd/ideas/...). If you have sensitive files or strict disk policies, review that location and permissions. 3) The spawned sub-agents perform web_search (network access) and may include external links/quoted data — treat outputs like any other web-researched content and validate sources. If any of these are concerns, restrict or review platform integrations or run the skill in an isolated environment.

Type: OpenClaw Skill Name: research-idea Version: 0.1.0 The skill is designed to launch a sub-agent for market research, utilizing `web_search` for data collection and saving results to a designated path (`~/clawd/ideas/<slug>/research.md`). Both network access (via `web_search`) and file system write access are explicitly stated and necessary for the skill's intended purpose. The instructions in `SKILL.md` and `templates/idea-exploration-prompt.md` are clear, well-structured, and do not contain any evidence of malicious intent, data exfiltration, unauthorized execution, or harmful prompt injection attempts. The quirky phrase in `README.md` is not an instruction to the agent and does not indicate malicious behavior.