$ loading_
Store agent memories cheaply and recall relevant context within strict token budgets.
Copy the install command and let the AI configure it · recommended for beginners
No copy-paste install info for "Thrift" yet — see the docs or source repo.
Store this conversation history and task log in Thrift, organize it into searchable memories by topic, and ensure future recalls stay within a 2000-token context budget.
Creates low-cost memory entries and returns the most relevant slices within the token limit.
I need to continue working on Client A's integration issue. Recall from Thrift the memories related to this client, error logs, and the previous solution, with a 1200-token budget, and include a recall receipt.
Returns the most relevant memory summaries or slices for the task, along with a recall receipt.
List the agent's last 10 memory recalls through Thrift, including the reason for recall, matched memory topics, token usage, and timestamps.
Produces an auditable recall log to review whether memory usage is efficient and compliant.
Manage AI agent memory cheaply with budgeted recall and savings monitoring.
Provide local-first semantic memory and context recall for MCP agents.
Give AI agents deterministic local project memory, context packs, and checkpoints.
Manage local-first agent memory with versioning, search, ACLs, and sync.
Give AI assistants persistent memory, adaptive recall, and graph-based knowledge retrieval.
Combine structured reasoning and execution so agents can think, act, and trace decisions.