Use this file to discover all available pages before exploring further.
The Model Context Protocol (MCP) is an open standard that lets AI assistants call tools exposed by external servers. Hindsight includes a built-in MCP server, so any MCP-compatible client — Claude Code, Claude Desktop, Cursor, or your own tooling — can store and retrieve memories without writing integration code.
The MCP server operates in two modes depending on the URL you connect to:
Mode
URL
bank_id
Single-bank (recommended)
/mcp/{bank_id}/
Implicit from URL — tools don’t expose a bank_id parameter
Multi-bank
/mcp/
Explicit parameter on each tool call — also exposes list_banks, create_bank, get_bank_stats
Single-bank mode is recommended for most use cases. It scopes all tool calls to the bank in the URL, which simplifies tool usage and enforces isolation per connection.
Generate thoughtful analysis by synthesizing stored memories with the bank’s personality and mission.
Parameter
Type
Required
Description
query
string
Yes
The question or topic to reflect on
context
string
No
Additional context about why this reflection is needed
budget
string
No
Search budget: low, mid, or high (default: low)
max_tokens
integer
No
Maximum tokens in the response (default: 4096)
response_schema
object
No
JSON Schema for structured output
tags
list[string]
No
Filter memories by tags before reflecting
{ "name": "reflect", "arguments": { "query": "Based on my past decisions, what architectural style do I prefer?", "budget": "mid", "tags": ["architecture"] }}
Use recall when you need raw facts (“What did Alice say about deadlines?”). Use reflect when you need reasoned analysis (“What should Alice prioritize this sprint?”).