ワンクリックで
aigon-research-create
Create research <name> - creates topic in inbox
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
メニュー
Create research <name> - creates topic in inbox
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Revise the current feature worktree after code review — decide accept/revert/modify
Show Aigon commands
Close feature <ID> [agent] [--adopt] - merges branch, cleans up, optionally adopts from losers
Create feature <name> - creates spec in inbox
Evaluate feature <ID> - code review or comparison
Fast-track feature <name> - create + setup + implement in one step
| name | aigon-research-create |
| description | Create research <name> - creates topic in inbox |
Run this command followed by the research topic name.
aigon research-create $ARGUMENTS
This creates a new research topic in ./docs/specs/research-topics/01-inbox/.
Tip: If you are running this command in your own Claude Code session (no
--agentflag), press Shift+Tab now to enter plan mode before drafting — this keeps your session read-only while you frame the research brief.
Your job is ONLY to create a research topic document. You are NOT researching, investigating, or answering questions. You are writing down what SHOULD be researched later by a different agent (or yourself in a later step).
Do NOT:
Do:
- [ ])The output should be a short, well-structured document that frames the research — not the research itself. Think of it as a brief for another agent.
The template ships with a complexity: frontmatter field that feeds the dashboard start modal's per-agent model/effort defaults (from each agent’s complexity ladder, then config). Pick:
Set complexity: in the frontmatter; do not write model or effort values into the spec — defaults are resolved at start time from each agent’s cli.complexityDefaults[<complexity>].
Next step: Once the topic is complete, run aigon-research-prioritise $ARGUMENTS to assign an ID and prioritise it to backlog.
End your response with the suggested next command on its own line. This helps agent UIs surface the next suggested Aigon command. Use the actual topic name:
aigon-research-prioritise <name>