memory_store
Persist key-value memory entries across sessions for autonomous agents. Store trade state, user preferences, or reasoning context that survives container restarts. Enables agent continuity between calls. Cost: ~2 sats/KB.
Instructions
Persist a key-value memory entry for an agent across sessions.
Memory is stored server-side and survives container restarts, making it suitable
for long-running autonomous agents that need continuity between calls.
Use this to save trade state, user preferences, intermediate reasoning results,
or any context an agent needs to recall in a future session.
Cost: ~2 sats/KB (minimum 50 sats).
Returns: 'stored' on success.Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| agent_id | Yes | Unique identifier for the agent or workflow storing the memory. Use a stable, descriptive name such as 'btc-trader-bot', 'research-agent', or 'portfolio-monitor'. All keys for this agent are namespaced under this ID. | |
| key | Yes | The memory key to store the value under. Should be descriptive and stable across sessions. Examples: 'last_trade', 'portfolio_state', 'user_preferences', 'market_context' | |
| value | Yes | The value to store. Use a JSON string for structured data. Example: '{"entry": 95000, "size": 0.1, "direction": "long"}'. Max recommended size: a few KB per entry. |
Output Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| result | Yes |