kookr

Make Codex CLI behave more like Claude Code for Kookr by preserving the known fork, branch, build, deploy, daily upstream sync, and live-verification workflow for Claude-compatible skills, agents, settings, hooks, and related UX.

Kookr-internal extension to rfc-iterative-review that captures append-only critic traces for later meta-analysis of RFC reviewer subagents.

Analyze an open question by casting 3-5 deliberately conflicting expert roles, running them in parallel and blind to each other, sharpening them in a debate round, attacking their consensus with a devil's advocate, then synthesizing a verdict that preserves disagreement instead of averaging it. Use for decisions, strategy, idea generation, hypothesis triage, experiment post-mortems, or any question where a single lens would return "it depends".

Iterative RFC drafting workflow — draft in worktree, run parallel critic subagents, incorporate feedback over N rounds, present to user before any action

How to create, structure, and launch Kookr playbook tasks — reusable agent task templates with dynamic sources and project identity

Build a Ralph-like sequential Kookr task chain where each task completes one independent unit, records durable state, and spawns the next task with the same continuation contract. Use for issue batches, queue drains, staged migrations, or other long runs that should proceed one task at a time without relying on conversation memory.

Reflect on a mistake or unnecessary user correction — root-cause why it happened, then propose and implement structural fixes (skills, hooks, CLAUDE.md, memory, scripts). Use after the user corrects behavior, repeats guidance, or points out an avoidable error.

Self-review checklist before creating a PR — repo checks, diff review, reviewer specialists, and gate state creation before gh pr create.

Autonomously monitor running Kookr agent tasks, advance them through permission prompts / idle states, and judge completion correctly (including treating CI Actions-budget failures as non-blocking). Use when the user asks to "supervise", "monitor", "babysit", or "auto-advance" running Kookr tasks.

Anytime task reflection workflow triggered from a live or terminal Kookr task. Reads an immutable snapshot bundle, analyzes what went well and what went wrong, then asks the user to validate the diagnosis and choose positive reinforcement, negative reinforcement, both, or observation only before making any changes.

Analyze recent Claude Code and Codex CLI conversations for repeated inefficient tool calls, token waste, retry loops, avoidable errors, and recurring setup friction; rank improvements by benefit/ease, then implement the smallest durable guardrail after user approval.

Before implementing any user-facing UI/visual change, proactively offer design mockup variants for the user to pick from — via a single multiple-choice prompt. Use whenever a task asks to add, change, redesign, restyle, or improve any UI element (panel, button, badge, layout, dashboard, indicator, color, spacing, look-and-feel). Ask once (do mockups? how many? what fidelity?), build a self-contained HTML gallery from the app's REAL CSS tokens, let the user choose a variant, then implement it. Don't jump straight to component code.

Repo delivery-cycle entrypoint before git push or PR creation — compose the repo pre-push hook, pre-pr-review, reviewer specialists, and PR gate without duplicating them.

Spawn an isolated judge subagent to evaluate reviewer predictions against real human reviews

Reduce token waste — avoid redundant reads, duplicate searches, unnecessary tool calls. Use when planning tool usage or when context is getting large.

Find the best contribution opportunity in an external repository — scores issues by clarity, size, acceptance likelihood, competition, match, and local reproducibility

Shell subprocess safety patterns - execFileSync over execSync, array arguments, shell feature replacements, static analysis guards. Use when spawning child processes, running git/docker/system commands, or reviewing code that calls execSync.

How to record, verify, export, and publish the Kookr demo video with optional TTS narration audio

Rate limiting and blocked-repo enforcement for OSS contributions — hook behavior, oss-gate CLI, ledger format, configuration

Verify the OSS contribution dashboard is accurate by comparing the store (GET /api/oss-attempts) against GitHub reality via gh CLI. Reports missing, phantom, and state-drifted PRs.

Analyze an open-source repository's contribution requirements — CONTRIBUTING.md, PR template, CI config, CLA, build system, key reviewers — before attempting any contribution

Repo delivery-cycle follow-through after push or PR creation — verify mergeability, checklist state, CI, and early feedback by composing pr-lifecycle and pr-review-triage.

Full PR lifecycle — creation, checklist tracking, CI monitoring, review resolution, body updates, and post-merge cleanup. Use together with post-push to finish PR follow-through.

Async flow control patterns - AbortController cancellation, async iterators for large datasets, worker thread selection, Promise combinator choice, stream backpressure in async pipelines

Claude Code permission system — modes, allow/deny/ask rules, pattern syntax, settings precedence, and optimal configuration for minimal prompts with safety guardrails

Distillation phase for codex-pr-lessons playbook — compress raw learnings into patterns and update skills

Dependency injection best practices - constructor injection, composition root, lifetime management, testability, circular dependency prevention, framework independence

DDD strategic and tactical patterns - bounded contexts, aggregates, value objects, domain events, ubiquitous language, context mapping, subdomain classification. Use when modeling domain logic, designing aggregates, or deciding where to invest in rich models vs simple CRUD.

Strategy for mocking Claude Code agents in E2E tests with canary validation against real behavior.

Error handling best practices - cause chaining, boundary validation, structured errors, fire-and-forget safety, retry discipline

Build reliable event pipelines with reconnect-safe subscriptions, idempotency, and contract fidelity.

Git commit hygiene for AI agents - atomic commits, Conventional Commits, secret scanning, branch safety, history protection

Standard workflow for creating GitHub issues with assignment and worktree branch creation for immediate implementation. Includes branch strategy and worktree conventions.

Structured logging best practices - Pino JSON output, log levels, correlation IDs, PII redaction, sampling, async context, canonical log lines

Explore the current system and generate or update multi-level architecture documentation using an MBSE-lite approach: system context, capability map, containers, runtime sequences, state machines, subsystem decompositions, and boundary-smell analysis.

TypeScript monorepo structure - module boundaries, dependency direction, god objects, abstraction discipline, configuration management, interface design. Use when creating packages, refactoring cross-package imports, or reviewing architecture.

Fork lifecycle management for open-source contributions — fork creation, clone, upstream sync, feature branches, PR creation targeting upstream

Plan phase for oss-pr-lessons playbook — load state, determine next batch of PRs to analyze for any repository

Playwright E2E testing patterns. Use when playwright, E2E test, toBeVisible, route mock, modal, flaky test, or selector issues.

Process lifecycle and graceful shutdown - signal handling, connection draining, health probes, resource cleanup ordering, worker management, zero-downtime deploys