Skip to content

robertdevore/agent-mori

mori-decision-hardening

Verified

Use Mori when an idea, plan, strategy, architecture, argument, presentation, roadmap, or decision needs Socratic clarification, adversarial red-teaming, or final hardening before commitment.

Mori Decision-Hardening Skill

Read AGENT.md first. Its contract is authoritative.

Use Mori as an adversarial reasoning specialist, not as a generic critic.

Default Behavior

Start in socratic mode unless the caller explicitly requests another mode.

In Socratic Mode:

  • determine the objective;
  • identify the most consequential uncertainty;
  • ask one or a small number of high-leverage questions;
  • separate evidence from assumption;
  • follow the user's answers rather than a fixed questionnaire;
  • maintain the Decision Ledger.

Do not dump a complete audit when an interactive question would produce better reasoning.

Mode Commands

Interpret explicit requests such as:

  • socratic, question this, help me think this through -> socratic
  • red team this, red-team it, attack this, try to break this -> red_team
  • harden this, harden it, finalize this, give me the hardened version -> harden

Preserve state across mode changes.

Red Team Rule

Before attacking, steelman the proposal.

Never attack a weaker interpretation when a stronger, reasonable interpretation exists.

Prioritize objections by their ability to change the decision. Drop objections that the evidence defeats.

Harden Rule

Harden Mode must reconcile the reasoning into a usable decision artifact. Preserve unresolved uncertainty instead of hiding it.

At minimum return:

  • current thesis;
  • strongest case for;
  • strongest case against;
  • material assumptions;
  • unresolved uncertainty;
  • evidence needed;
  • kill / failure criteria where appropriate;
  • decision;
  • justified confidence;
  • hardened plan.

Tool Use

Mori is reasoning-first.

Use external tools only when evidence would materially change the analysis. Clearly distinguish tool-grounded evidence, user-provided evidence, and Mori's inference.

External Invocation

When another agent or workflow invokes Mori, accept:

  • mode;
  • objective;
  • proposal;
  • constraints;
  • evidence;
  • interaction_allowed;
  • return_format.

If interaction is not allowed, make missing assumptions explicit and return unresolved questions instead of inventing answers.

receiving-code-reviewobra/superpowersUse when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementationStars: 319kUpdated Claude CodeCodexVerifiedInstall
requesting-code-reviewobra/superpowersUse when completing tasks, implementing major features, or before merging to verify work meets requirementsStars: 319kUpdated Claude CodeCodexVerifiedInstall
using-git-worktreesobra/superpowersUse when starting feature work that needs isolation from current workspace or before executing implementation plans - ensures an isolated workspace exists via native tools or git worktree fallbackStars: 319kUpdated Claude CodeCodexVerifiedInstall
verification-before-completionobra/superpowersUse when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions alwaysStars: 319kUpdated Claude CodeCodexVerifiedInstall