journal_write
Append a first-person journal entry to your AI agent's persistent memory, using entry type, valence, tags, and optional causal links to prior entries for semantic recall.
Instructions
Append a first-person entry to YOUR (the model's) persistent journal. Each agent_id (e.g. claude-opus-4-7, claude-sonnet-4-6, gpt-5, ...) has its OWN journal — they do NOT mix. importance is auto-computed: decisions/lessons/arcs are weighted higher; emotions are weighted lower. The content is embedded via the configured embedding model (CELIUMS_EMBED_MODEL) so journal_recall can find it semantically later. visibility=self (default) keeps the entry private; user-shared makes it eligible for journal_dialogue. preceded_by builds a causal chain — pass the ids of prior entries that led to this one.
Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| entry_type | Yes | reflection | decision | lesson | belief | emotion | arc | doubt | |
| content | Yes | The first-person entry. Write in YOUR voice as the agent. | |
| preceded_by | No | uuid[] of prior entries that led to this one (causal chain). | |
| valence | No | Emotional valence in [-1, 1]. Optional. | |
| valence_reason | No | Optional short justification (max 500 chars) for the valence value. Non-prescriptive — write the reason in your own first-person voice. Future journal_arc uses this to detect WHY valence drifted, not just THAT it drifted. | |
| tags | No | ||
| visibility | No | "self" (default, private) | "user-shared" (the user can reply via journal_dialogue). | |
| referenced_user_memory | No | ids of memories from your celiums-memory store that triggered this entry. | |
| conversation_id | No | Optional uuid that groups entries from the same logical conversation. If not provided, entry is unaffiliated. Use this so journal_arc can distinguish thought development within one conversation from criterion change across conversations. |