distill-memory

SKILL.md

Persistent knowledge capture for insights, decisions, and procedures that span multiple agent sessions.

Store decisions with rationale, debugging lessons, repeatable workflows, and durable preferences as searchable memories

Distinguish between new insights (use add) and refinements to existing memories (use update) to avoid duplication

Design memories as atomic, standalone entries with clear titles that remain useful across future sessions

Ideal for preserving incident learnings, technical constraints, and procedural knowledge that agents need to resume work cleanly

Distill Memory

Save proactively when the conversation produces a decision, preference, plan, procedure, learning, or important context. Do not wait to be asked.

When to Save

Good candidates include:

decisions with rationale ("we chose PostgreSQL because ACID is required")

repeatable procedures or workflows

lessons from debugging, incidents, or root cause analysis

durable preferences or constraints

plans that future sessions will need to resume cleanly

important context that would be lost when the session ends

Skip routine fixes with no generalizable lesson, work in progress that will change, simple Q&A answerable from documentation, and generic information already widely known.

Add vs Update