From predictive-mind
Given a goal and a set of candidate next actions, rank them by Expected Free Energy (pragmatic + epistemic value) using past similar episodes, and pick the winner. Delegates scoring to the policy-selector subagent.
How this skill is triggered — by the user, by Claude, or both
Slash command
/predictive-mind:plan-efeThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Upholds Invariant 3 of predictive-mind: **pragmatic and epistemic value
Upholds Invariant 3 of predictive-mind: pragmatic and epistemic value are scored separately. Use this skill when multiple candidate actions are on the table and the "first plausible one" is not obviously the best.
/probe, /reflect, or a new action./probe call that would disambiguate an open
hypothesis).task_id (so the subagent can call retrieve_beliefs).ask-user, stop and ask. If the pick is
probe-first, invoke /probe. Otherwise proceed with
/predict then the chosen action.recall_similar_episodes for each candidate, so prior surprises
actually discount similar-looking plans.$ARGUMENTS — the current goal (free text). The skill will ask you
for the candidate list if you did not include it.
ranked: [ { action, pragmatic, epistemic, total, recommendation, ... } ]
pick: { action, rationale }
notes_on_calibration: <honest caveats>
Use the pick's recommendation field to decide the next skill:
execute → /predict then run the actionprobe-first → /probedream-first → invoke the dreamer on the picked action before actingskip → drop this candidate, re-plannpx claudepluginhub jason-hchsieh/predictive-mind --plugin predictive-mindGuides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.