Accreted Intelligence Skills

Use a local work model to preserve commitments, task frames, and outcome-based learning.

AccInt’s complete reference-host Skill package contains commitments, frames, and solve workflows. The underlying local MCP daemon and retrieval model preserve work context and learn from real outcomes.

Use cases

  • Frame and solve a continuing task
  • Track commitments across sessions
  • Retain lessons from task outcomes

How to use

Install the upstream AccInt daemon and connect the reference Claude host. Prepare the hardware and storage for its selected embedding model; keep the local work state available between sessions.

Skills

  • commitments

    Triage acc's open promises and close them with honest real-world verdicts via acc_act(runtime="outcome").

    maxbaluev/accreted-intelligence
  • frames

    Drain acc's deliberation queue — open/waiting brain_frames checkpointed by headless runs — via acc_act(runtime="continue").

    maxbaluev/accreted-intelligence
  • solve

    Route a goal through acc's scored-memory loop via acc_act(runtime="solve"); deliberate any returned brain_frame and submit via continue.

    maxbaluev/accreted-intelligence