agent-intel — loop distillation, skill RAG, predictive routing

Package scope and prerequisites

This catalog ships instructions, not the executable runtime. Python modules, scripts, local services and optional tools named below must be installed separately from the upstream project. Check their availability before running commands; if they are missing, explain the prerequisite instead of claiming execution or results. The host agent may use model tokens while following these instructions.

Commands, APIs and runtime behavior described below are conditional examples for a compatible upstream installation; they are not evidence of execution in this package.

Purpose

Aggregates what individual runs learned: loops are recorded and distilled, skills are selected for a task by retrieval, fix cost and agent route are predicted from history, and agent memory is compressed.

python -m skills.agent_intel.cli loop-distill
python -m skills.agent_intel.cli skill-rag
python -m skills.agent_intel.cli predict-fix
python -m skills.agent_intel.cli predict-route
python -m skills.agent_intel.cli distill
python -m skills.agent_intel.cli stats

When NOT to use

• You want failure patterns turned into correction rules — use botte_learn.
• You want to send a task to local hardware — use local_router.
• You want a tool chosen for one call — use tool_router or skill_finder.