From career-copilot
The self-learning brain of Career Copilot. Reads outcome history across the pipeline, computes what's actually working (reply rate by message pattern, conversion by fit tier and framing, referral vs cold), and rewrites the active strategy playbook so future research/tailor/outreach runs reason better. Use when the user says "what's working", "update my strategy", "coach me", "learn from my results", after recording outcomes, or on a weekly schedule.
How this skill is triggered — by the user, by Claude, or both
Slash command
/career-copilot:coachThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
This is the meta-loop. Every other skill acts and records; `coach` reads the record and
This is the meta-loop. Every other skill acts and records; coach reads the record and
rewrites the playbook the others read.
outcomes.md event logs across the pipeline (+ each record's stage/status).~/Career Copilot/strategy.md (create it if absent).references/metrics.md):
~/Career Copilot/strategy.md — a short, current playbook: what
to do more of, what to stop, target adjustments, and the default outreach pattern. This is
exactly what the SessionStart hook injects into every future session.strategy.md is a living one-pager, not an archive.research / tailor-resume / outreach
run is measurably better because of what was learned.npx claudepluginhub aniebyl/career-copilot --plugin career-copilotCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.