Skill Claude CodeCodex
Deep security audit — OWASP ASVS 5.0 + Top 10 (+ LLM/Agentic when AI-bearing), STRIDE threat model, SAST/SCA/secret scanners as evidence; freeze to {runs}/security/security-NN.md. Non-advancing. Use before ship on any UI/API/data surface.
Skill Claude CodeCodex
Read-only lifecycle status — reads the state file (paths.state) and prints the current phase, its gate status, and the single next action/command. Cheap; run anytime to orient or resume.
Skill Claude CodeCodex
Standing adversarial debate — Defense vs Prosecution + Catfish across idea, market, arch, scope, rules, and code; orchestrator judges per claim and freezes {runs}/debates/debate-NN.md. Use on demand or before big gates (pre-go/no-go, pre-rules-freeze, pre-ship); not a sprint gate.
Skill Claude CodeCodex
Sprint UI/API verification — drive flows, inspect runtime health, freeze per-claim verdicts to {runs}/verifications/verify-NN.md. Use after a UI-touching sprint lands, before /close-sprint; hard-skips non-UI sprints.
Skill Claude CodeCodex
Phase 6 — decompose the MVP scope into a dependency-ordered roadmap and per-sprint plans, each with goal, scope, tasks, acceptance criteria, and a DoD that references the frozen rules. Use after conventions and the design system are frozen (stage architecturerules → sprintplanning), before any sprint executes.
Skill Claude CodeCodex
Phase 7 kickoff — pre-sprint drift audit vs frozen rules, working plan, set sprint active. Use to begin a planned sprint (stage sprintplanning → sprintexecution). Do NOT use to close a sprint or grade conformance — that is /close-sprint after tasks, tests, and /midas-verify.
Agent Claude Code
Delegate here to IMPLEMENT and WRITE ARTIFACTS — Phase 7 code + tests, and writing docs, ADRs, market/business/architecture documents, rules, and sprint files. The build tier; produces the on-disk artifacts the orchestrator later audits.
Agent Claude Code
Delegate here for the irreversible THINK / PLAN / AUDIT / DECIDE decisions — Phase 1 gap loops, Phase 4 stack choice, Phase 5 rule design, Phase 6 sequencing, Phase 8 conformance audits, and any exit-gate verdict. The strongest tier; reserved for decisions that are costly to undo.
Agent Claude Code
Delegate here for fast, mechanical SEARCH / EXTRACT work — file and status extraction, Context7 doc fetches, competitor/source gathering, and evidence collection for an audit. The cheapest tier; read-only, returns facts and exact citations, never decides.
Settings file Claude Code
Agent settings declaring 5 allowed tools.
Cursor rule Cursor
Midas base conventions (always-on project law). Generated from harness/conventions.md.
Cursor rule Cursor
Midas CHECK digest for Phase-8 conformance audits and rule reviews. Full always-on rule CHECKs (base + project overlays). Attach when auditing, closing a sprint, or verifying rule compliance.
Instructions file
Instructions for okuzpe/midas-harness, a project described as: Midas — a portable product-development harness for AI agents: drive a product from idea to shipped code through 9 audited phases, with cost-tiered agents, Context7, and a whole-project adversarial debate (/midas-tribunal).
Instructions file Gemini CLI
Instructions for okuzpe/midas-harness, covering project memory — midas (gemini cli), midas base conventions (always-on), precedence (when rules conflict, higher wins), path resolution and code quality.
Skill Claude CodeCodex
Run diff-scoped test/quality gate receipts into {paths.cache}/gates/ /; use before close-sprint when production paths changed. Does not replace close-sprint or midas-verify.
Skill Claude CodeCodex
Over-engineering review — delete-list for the current diff or named paths (stdlib/native/yagni/shrink). Use before /close-sprint on a fat diff, after a feature branch, or when the user asks what to cut. Complements correctness review; does not replace /midas-sweep or /close-sprint.
Skill Claude CodeCodex
Phase 7 STM writer — updates {runs}/sprints/NN-progress.md after tasks or significant decisions (Done rows, Learned observations, Next line). Build-tier; complements read-only /midas-recall. Use mid-sprint when session-continuity rule applies.
Skill Claude CodeCodex
Ad-hoc branch/PR QA — map diff to routes/screens, drive with agent-browser (web) or Maestro MCP (native), report in chat; optional {runs}/qa/qa-adhoc-.md. Use during Phase 7 inner loop; does not replace /midas-verify before /close-sprint.
Skill Claude CodeCodex
Hygiene and dead-flow sweep — find orphans, stale docs, ledger drift, and harness mismatches; freeze to {runs}/sweeps/sweep-NN.md. Optional --fix with explicit confirm. Use on demand, after brownfield adopt, or before closing a large sprint.
Plugin Claude Code
Marketplace for the Midas product-development harness.
Plugin Claude Code
Portable product-development harness: drive a product from idea to shipped code through 9 audited phases, with cost-tiered agents, a Context7-first rule, and a whole-project adversarial debate (/midas-tribunal).