/ask responses should not prompt for follow-up interaction
- Langage dominant
- Shell
- Étoiles
- 11.2k
- Forks
- 1.9k
- Merge moyen
- 14 h 16 min
- PR mergées (30 j)
- 6
Description
### Describe the feature or problem you'd like to solve
/ask should not prompt for replies because the user cannot interact in a multiturn way in this side session
### Proposed solution
Description:
/ask is documented as a single-turn side question that doesn't add to conversation history. However, the model's responses in /ask sessions often propose next steps, ask follow-up questions, and prompt the user to reply — even though there's no way to interact further. The
only option is pressing Esc, which discards the entire context.
Current behavior:
- User runs /ask with a question
- Model responds with an answer, then says something like "Would you like me to dig deeper into X?" or "Here are some options — which would you prefer?"
- User has no way to respond — can only press Esc and lose the context
Expected behavior:
- Model responses in /ask should be self-contained — give the full answer without expecting follow-up
- If there are natural next steps, frame them as actions the user can take back in the main session (e.g., "To explore this further, try asking in your main session: ..."). Provide something you can copy paste from this side conversation into your main or new session to take the next steps.
- Never ask questions or prompt for replies in /ask mode
Impact: Creates a confusing dead-end experience. Users feel like they're missing functionality or hitting a bug, when it's actually the model not adapting its response style to the single-turn constraint.
### Example prompts or workflows
_No response_
### Additional context
_No response_
Guide de contribution
Ouvrir le guide de contribution
Piste de recherche
Commencez au point d’entrée /ask de la CLI et suivez la manière dont son contexte en un seul tour est envoyé au modèle ; l’issue n’identifie aucun fichier ni test. Le travail est terminé lorsque les réponses de /ask sont autonomes, ne demandent jamais de réponse et fournissent des prompts ou des actions copiables pour poursuivre dans la session principale.
Rédigé par le modèle d'indexation à partir du texte de l'issue.
Évaluation
- Stack technique
- shell
- Domaine
- ai, cli
- Type d'issue
- Fonctionnalité
- Difficulté
- 4/5
- Temps estimé
- 3-5 jours
- Activité
- Calme
- Clarté
- Plutôt claire
- Accessibilité débutants
- 38/100