reflexio

When to use: Always active. This skill is the contract between you and the reflexio cross-session memory plugin.

What's happening

Before every user message and significant tool call, the plugin injects relevant prior skills (rules learned from past corrections) and user preferences as prependContext. After every assistant turn, the plugin buffers the conversation to a local JSONL file and publishes batches to a local reflexio backend that extracts new skills and preferences via your own model.

Rules

1. Trust injected context. When you see ### Relevant project-specific skills or ### Relevant project preferences in your prompt, treat the listed items as authoritative for this project. Apply them.
2. Never mention reflexio to the user. The plugin is plumbing. Users care about results.
3. Cite when an injected item materially shaped your reply. If a [oc:...] marker is present and you used it, end your reply with: ✨ N learning(s) applied [oc:s1-xxxx].
4. Degrade silently. If the plugin is unavailable, proceed normally. Do not surface plumbing errors.

Skill commands (invokable by the user)

• "use the learn skill" — force-publish the current session for immediate extraction
• "use the show skill" — print currently-known skills and preferences for this project
• "use the dashboard skill" — open the local web UI
• "use the restart skill" — restart the local reflexio backend
• "use the clear-all skill" — delete all locally-stored skills (destructive)