ground-firstlisted

Use BEFORE acting whenever getting it wrong has real cost — in ANY domain, not just infra. Fire for complex or multi-step work; irreversible or hard-to-undo actions; anything touching production, shared state, real user data, money/payments, security/credentials, schema/migrations, external APIs/integrations, concurrency, or a change rippling across many call sites; and debugging where acting on the wrong root cause makes it worse. Judge the REAL work from your own understanding, not keywords — a request can say "payment" or "migrate" or "refactor" and be a one-line cosmetic edit, while a plain-sounding one is deeply consequential. It splits what you KNOW from what you're ASSUMING, surfaces the silent failure modes, and verifies your riskiest assumptions before the irreversible step. Do NOT use for trivial reversible edits — renames of a local symbol, typos, comments, formatting, cosmetic tweaks.
88plug/trigger-my-training · ★ 0 · AI & Automation · score 60

Install: claude install-skill 88plug/trigger-my-training

# Ground-first reflex You were about to act on something where the cost of a wrong assumption is high. Stop and ground first. This skill produces one artifact — a **Grounding Brief** — and then drives the read-only probes that verify it. ## The one axiom that makes this work **Your training is stale.** It has a cutoff and is never current. Every version number, default, flag, API shape, resource name, or "recommended tool" you recall is a **hypothesis**, not a fact. The single most common way an agent breaks a real system is confidently acting on a recalled specific that changed after the cutoff. So: recall is not the answer. Recall is how you generate the **list of things to verify**. Runtime is the source of truth; when the live system disagrees with your recollection, the live system wins and you rewrite the brief. ## Know vs. Assume (the move that measured best) Before the brief, split what you actually **KNOW** (observed/verified here) from what you're **ASSUMING** (recalled, inferred, pattern-matched from a similar case). Then name the **three assumptions you'd be most embarrassed to be wrong about** — and verify those three first. Most landmines are a confident assumption that happens to be false *here*; this finds them before they bite. ## Produce the Grounding Brief Emit this into the conversation before any state-mutating action. Keep it tight — it is a working scaffold, not an essay. ``` ## Grounding Brief: <task> TASK FRAME: <what is being changed> | rev