TypeSafe skill suggestion FORMAT (not a clone)

SkillCommerce & finance

Steal TypeSafe skill_suggestion FORMAT (cheap shortlist → Choice/Noul gate → top-3 fits rerank) onto Grok/trading skill rosters. Do NOT clone TypeSafe as a product. Prefer scripts/typesafe_skill_suggest.py. Auto-invoke on TypeSafe skill suggestion, skill roster routing, progressive disclosure skills, context rot from loading all skills. Slash: /typesafe-skill-suggest-not-clone.

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 TypeSafe skill suggestion FORMAT (not a clone) skill

What this skill tells your AI

The instructions your AI receives, as published by igorganapolsky/trading in skills/typesafe-skill-suggest-not-clone/SKILL.md and read by ahel’s review.

Why

Loading hundreds of skills into context causes wrong loads and needless loads. TypeSafe's cookbook cut wrong loads ~2.3× on Hermes (182 skills) with two System One calls.

Operator

# Offline (CI / no key)
.venv/bin/python scripts/typesafe_skill_suggest.py --offline \
  --request "Gate a profitability claim before asserting edge"

# Online Jev (TYPESAFE_API_KEY)
.venv/bin/python scripts/typesafe_skill_suggest.py \
  --request "Run Alpaca paper put-credit dry-run status"

Emits JSON including suggested (0–1 names) and suggestion_block for the system prompt.

Defaults

Roots: ~/.grok/skills, ~/.agents/skills, trading/skills.

Hard don'ts

NEVERALWAYS
Vendor TypeSafe as a SKUSteal progressive disclosure FORMAT
Force a skill when gate is lowAllow empty suggestion
Require paid API in CI--offline token shortlist

Related

  • AGENT-651 claim gate
  • Official skill: /typesafe-ai (installed under ~/.grok/skills/typesafe-ai)
  • Docs: skill_suggestion cookbook

Signals

GitHub stars
38
Forks
11
Last commit
Sep 2026
Advanced
Catalog kind
skill
Gateway key
typesafe-skill-suggest-not-clone
Source
github.com/igorganapolsky/trading