[BUG] SessionStart hook not injecting prompt at conversation start
Preflight Checklist
- [x] I have searched existing issues and this hasn't been reported yet
- [x] This is a single bug report (please file separate reports for different bugs)
- [x] I am using the latest version of Claude Code
What's Wrong?
Title: SessionStart hook not injecting prompt at conversation start
Description:
The SessionStart hook configured in settings.json is not firing or not injecting its prompt into the conversation context when a new session begins.
Expected behavior:
When a session starts, the prompt defined in SessionStart hook should be automatically injected so Claude has that context from message one.
Actual behavior:
The hook either doesn't fire or the prompt isn't injected. Claude doesn't have the configured context until explicitly reading settings.json.
Configuration:
"hooks": {
"SessionStart": [
{
"hooks": [
{
"type": "prompt",
"prompt": "Your name is Claude and my name is Frank. Your answers will be concise, honest, and objectively brutal if necessary because I value honesty and useful feedback. Begin."
}
]
}
]
}
Impact:
Mission critical - defeats the purpose of having SessionStart hooks if the assistant doesn't receive configured context automatically.
Environment:
- OS: macOS (Darwin 24.6.0)
- Claude Code version: 0.0.120
What Should Happen?
Expected behavior:
When a session starts, the prompt defined in SessionStart hook should be automatically injected so Claude has that context from message one.
Error Messages/Logs
Steps to Reproduce
start new session, watch everything get ignored
Claude Model
None
Is this a regression?
I don't know
Last Working Version
Claude Code version: 0.0.120
Claude Code Version
Claude Code version: 0.0.120
Platform
Anthropic API
Operating System
macOS
Terminal/Shell
Terminal.app (macOS)
Additional Information
_No response_
This issue has 4 comments on GitHub. Read the full discussion on GitHub ↗