engram-agent 0.2.1

Reusable LLM agent loop with tool-calling and lifecycle hooks
docs.rs failed to build engram-agent-0.2.1
Please check the build logs for more information.
See Builds for ideas on how to fix a failed build, or Metadata for how to configure docs.rs builds.
If you believe this is docs.rs' fault, open an issue.

engram-agent: Reusable LLM agent loop with tool-calling and lifecycle hooks.

This crate provides a generic agent loop that:

  • Calls an LLM with tool schemas
  • Dispatches tool calls to [Tool] implementations
  • Applies lifecycle hooks via [AgentHook]
  • Handles duplicate detection, cost limits, and iteration limits

Architecture

engram-core (owns LlmClient trait + HttpLlmClient impl)
    ↑
engram-agent (owns Agent loop + Tool/AgentHook traits)
    ↑
engram-bench (implements Tool for memory tools, AgentHook for gates)
engram-server (future: implements Tool for REST-backed tools)