Reflect

SkillDocs & knowledge

Extract durable, non-obvious lessons from the current dialogue and route them to project guidance, memory, documentation, or the active task conclusion. Use when the user asks to reflect on a session or preserve learnings.

Available today. Use it from your connected AI after setup.

Connect ahel once, and every AI you use reads what you have installed.

Then ask your AI: use the Reflect skill

What this skill tells your AI

The instructions your AI receives, as published by btseytlin/ultrapack in skills/reflect/SKILL.md and read by ahel’s review.

Use this as Codex's equivalent of /up:reflect.

  1. Identify non-obvious user corrections, discovered conventions, external-system details, and decisions the code does not explain. Discard one-off debugging facts and anything already documented.
  2. Route each remaining lesson to the right home: CLAUDE.md for durable repository guidance, memory for cross-session personal context, project docs for domain knowledge, or the task conclusion for task-specific discoveries.
  3. Write concise actionable entries: what the rule is, why it matters when non-obvious, and when it applies. Do not duplicate a learning across destinations.
  4. Show the user a proposed diff before committing changes to CLAUDE.md. Re-read every entry as a fresh agent would and tighten anything that still needs unstated context.

Signals

GitHub stars
47
Forks
12
Last commit
Sep 2026
Advanced
Catalog kind
skill
Gateway key
reflect-btseytlin
Source
github.com/btseytlin/ultrapack