loop▌

Automated agent-driven development loop that executes AI agents against user story acceptance criteria. \n \n Structures features as JSON-formatted user stories with testable acceptance criteria that agents can verify and track \n Runs AI agents in a continuous loop to implement features, check acceptance criteria, and log progress for subsequent agent iterations \n Requires prerequisite setup of AI coding agent configuration and user stories framework before running the Ralph agent loop \n Inte

Decision framework for autonomous agent loops with quality gates, evals, and recovery patterns. \n \n Four loop modes available: continuous-pr for CI/PR workflows, rfc-dag for RFC decomposition, infinite for exploratory parallel generation, and sequential as default \n Recommended production stack combines RFC decomposition, code quality gates, eval harness, and session persistence for robust autonomous execution \n Built-in failure mode detection covers loop churn, repeated retries, merge queue

Use this skill to run and test the app locally with Docker Compose, then hand off to Eve for staging deploys.

Start a recurring experiment loop that runs at a user-selected interval.

Comprehensive verification system running build, type, lint, test, and security checks across Claude Code sessions. \n \n Executes six sequential verification phases: build compilation, type checking, linting, test suite with coverage reporting, security scanning for secrets and debug statements, and git diff review \n Supports both JavaScript/TypeScript (npm/pnpm, tsc, ESLint) and Python (pyright, ruff) projects with language-specific commands \n Produces a structured verification report indica

Skill by ara.so — Daily 2026 Skills collection.

You are setting up a loop — an iterative worker/verifier cycle managed by the Paseo daemon.

$22

Pause graph execution for human review, approval, or validation, then resume with their input. \n \n Requires three components: a checkpointer (InMemorySaver or PostgresSaver), a thread ID in config, and JSON-serializable interrupt payloads \n interrupt(value) pauses and surfaces data; Command(resume=value) resumes and returns that value to the paused node \n All code before interrupt() re-executes on resume, so side effects must be idempotent (use upsert, not insert) \n Supports approval workfl