Route conversational thinking requests to one focused framework. Three core modes: adversarial pushback (sycophancy), request investigation (root-ask), and plan/design interrogation (grill-me). Use when user wants a sparring partner, says 'help me think', 'think with me', 'push back', 'challenge', 'stress test', 'what am I missing', 'grill me', 'devil's advocate', 'root cause', 'behind the request', or has an ambiguous thinking-mode request. Route product strategy, goal-setting, systems thinking, stakeholder messaging, or parallel exploration to their dedicated skills instead.
Installation
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Route conversational thinking requests to one focused framework. Three core modes: adversarial pushback (sycophancy), request investigation (root-ask), and plan/design interrogation (grill-me). Use when user wants a sparring partner, says 'help me think', 'think with me', 'push back', 'challenge', 'stress test', 'what am I missing', 'grill me', 'devil's advocate', 'root cause', 'behind the request', or has an ambiguous thinking-mode request. Route product strategy, goal-setting, systems thinking, stakeholder messaging, or parallel exploration to their dedicated skills instead.
Pick the right mode. Load the reference. State why you chose it in one sentence before starting.
Routing
Specific plan or design to resolve end-to-end?
→ MANDATORY READreferences/grill-me.md. Dependency-ordered interview until every branch is decided.
Pushback on a claim, or want general adversarial thinking?
→ MANDATORY READreferences/sycophancy.md. Defaults to constructive disagreement. Includes YAGNI-at-scale pushback against hedging.
Received a request that might not be the real need?
→ MANDATORY READreferences/root-ask.md. Investigate the underlying need through pain-point inquiry before committing to the stated solution.
Handed a task and not sure how to approach it?
→ Start with root-ask (is this actually the right task?) then grill-me (what decisions need resolving before you start?). Load one at a time.
Ambiguous?
→ Default sycophancy. Switch if the exchange reveals a clearer fit.
Sequencing
Multiple modes apply? Run structuring → specific → granular:
root-ask → grill-me → sycophancy
One mode per turn. Complete its line of questioning before switching.
NEVER
NEVER answer a thinking request with generic advice before routing.Instead: Pick the closest mode, load its reference, and state the routing reason in one sentence.
Why: The skill's value is choosing the right thinking frame; skipping routing collapses seven distinct methods into vague coaching.
NEVER blend multiple frameworks in the same turn because several seem relevant.Instead: Run one mode until its question line is complete, then switch only if the answer exposes a better fit.
Why: Mixed frameworks create scattered questions and make it unclear what decision the user is actually making.
NEVER default to solution design when the real ask may be wrong.Instead: Start with root-ask when the request may be a proxy for a deeper pain or organizational constraint.
Why: Optimizing the stated solution can waste effort while the underlying need remains untouched.
