AI Agent Workflow Collection

Track and optimize agent specialization during methodology development. Use when agent specialization emerges (generic agents show >5x performance gap), multi-experiment comparison needed, or methodology transferability analysis required. Captures agent set evolution (An tracking), meta-agent evolution (Mn tracking), specialization decisions (when/why to create specialized agents), and reusability assessment (universal vs domain-specific vs task-specific). Enables systematic cross-experiment learning and optimized M0 evolution. 2-3 hours overhead per experiment.

One-time initializer for the backlog task board. Checks that the backlog CLI is installed, initializes a backlog project if none exists, and verifies that all columns required by loop-backlog and feature-to-backlog are present. Idempotent — safe to run multiple times.

Achieve comprehensive baseline (Vmeta ≥0.40) in iteration 0 to enable rapid convergence. Use when planning iteration 0 time allocation, domain has established practices to reference, rich historical data exists for immediate quantification, or targeting 3-4 iteration convergence. Provides 4 quality levels (minimal/basic/comprehensive/exceptional), component-by-component Vmeta calculation guide, and 3 strategies for comprehensive baseline (leverage prior art, quantify baseline, domain universality analysis). 40-50% iteration reduction when Vmeta(s0) ≥0.40 vs <0.20. Spend 3-4 extra hours in iteration 0, save 3-6 hours overall.

Comprehensive CI/CD pipeline methodology with quality gates, release automation, smoke testing, observability, and performance tracking. Use when setting up CI/CD from scratch, build time over 5 minutes, no automated quality gates, manual release process, lack of pipeline observability, or broken releases reaching production. Provides 5 quality gate categories (coverage threshold 75-80%, lint blocking, CHANGELOG validation, build verification, test pass rate), release automation with conventional commits and automatic CHANGELOG generation, 25 smoke tests across execution/consistency/structure categories, CI observability with metrics tracking and regression detection, performance optimization including native-only testing for Go cross-compilation. Validated in meta-cc with 91.7% pattern validation rate (11/12 patterns), 2.5-3.5x estimated speedup, GitHub Actions native with 70-80% transferability to GitLab CI and Jenkins.

Systematic methodology for standardizing cross-cutting concerns (error handling, logging, configuration) through pattern extraction, convention definition, automated enforcement, and CI integration. Use when codebase has inconsistent error handling, ad-hoc logging, scattered configuration, need automated compliance enforcement, or preparing for team scaling. Provides 5 universal principles (detect before standardize, prioritize by value, infrastructure enables scale, context is king, automate enforcement), file tier prioritization framework (ROI-based classification), pattern extraction workflow, convention selection process, linter development guide. Validated with 60-75% faster error diagnosis (rich context), 16.7x ROI for high-value files, 80-90% transferability across languages (Go, Python, JavaScript, Rust). Three concerns addressed: error handling (sentinel errors, context preservation, wrapping), logging (structured logging, log levels), configuration (centralized config, validation, environment variables).

Keep project documentation accurate when behavior changes. EXCLUSIVE to project-manager agent.

Interactive epic authoring for the B′′ epic lane. Drives a brief multi-child epic idea through propose then plan, each gated by human-reviewed cycles, and parks it at Epic: Backlog. The plan enumerates intended child sub-tasks but creates none — the human promotes Epic: Backlog to Epic: Ready to authorize autonomous decomposition by the epic worker.

Executes the full feature development lifecycle (phases 3-9): generates proposal documents from discussion, iteratively reviews and refines proposal/plan documents from a strict architect's perspective, then implements the plan using TDD with parallel Task agents in worktrees, and validates with self-analysis. Use when given a feature discussion or existing proposal/plan documents to generate, review, refine, and implement.

Drives the Proposal-to-Backlog pipeline for a software feature: two architect-reviewed cycles (one for Proposal, one for Plan), each converging on APPROVED within 8 rounds. Outputs a planSet entry ready for implementation; does not create branches or open PRs.

Progressive learning methodology for structured onboarding using time-boxed learning paths (Day-1, Week-1, Month-1), validation checkpoints, and scaffolding principles. Use when onboarding new contributors, reducing ramp-up time from weeks to days, creating self-service learning paths, systematizing ad-hoc knowledge sharing, or building institutional knowledge preservation. Provides 3 learning path templates (Day-1: 4-8h setup→contribution, Week-1: 20-40h architecture→feature, Month-1: 40-160h expertise→mentoring), progressive disclosure pattern, validation checkpoint principle, module mastery best practice. Validated with 3-8x onboarding speedup (structured vs. unstructured), 95%+ transferability to any software project (Go, Rust, Python, TypeScript). Learning theory principles applied: progressive disclosure, scaffolding, validation checkpoints, time-boxing.

Autonomous unified B′′ Worker for the backlog.md board. Drives BOTH lanes from one Monitor session: basic-ready events execute kind:basic tasks in isolated git worktrees that merge back on success; epic-ready events auto-decompose a kind:epic task (Epic: Ready → Decomposing → children at Basic: Backlog → Awaiting Children); child-done events reconcile the parent epic and, when all children are Basic: Done, run evaluate and write a FINISH/ITERATE recommendation for human confirmation. Invoke /loop-backlog once; it runs until the backlog/.loop-stop sentinel is written.

Apply Bootstrapped AI Methodology Engineering (BAIME) to develop project-specific methodologies through systematic Observe-Codify-Automate cycles with dual-layer value functions (instance quality + methodology quality). Use when creating testing strategies, CI/CD pipelines, error handling patterns, observability systems, or any reusable development methodology. Provides structured framework with convergence criteria, agent coordination, and empirical validation. Validated in 8 experiments with 100% success rate, 4.9 avg iterations, 10-50x speedup vs ad-hoc. Works for testing, CI/CD, error recovery, dependency management, documentation systems, knowledge transfer, technical debt, cross-cutting concerns.

Generate ready-to-use, context-aware prompts for the most natural next steps based on recent conversation activity. Pure reasoning skill — no MCP tool calls. Infers completion state, suggests primary continuation, and offers 2-3 ranked alternatives.

Achieve 3-4 iteration methodology convergence (vs standard 5-7) when clear baseline metrics exist, domain scope is focused, and direct validation is possible. Use when you have Vmeta baseline ≥0.40, quantifiable success criteria, retrospective validation data, and generic agents are sufficient. Enables 40-60% time reduction (10-15 hours vs 20-30 hours) without sacrificing quality. Prediction model helps estimate iteration count during experiment planning. Validated in error recovery (3 iterations, 10 hours, Vinstance=0.83, Vmeta=0.85).

Run a pre-registered quantitative experiment with held-out fixtures, multi-model k=5 execution, statistical verdict, and evidence write-back. Use when starting a new domain experiment that requires hard verdicts (CONFIRMED/NULL/REJECTED) rather than soft self-assessment.

Systematic methodology for constructing compact (<150 lines), expressive, Claude Code-integrated subagent prompts using lambda contracts and symbolic logic.

Creates a Ready-status backlog task from a pre-approved template, bypassing the full review cycle.

Creates a Ready-status backlog task from a pre-approved template, bypassing the full review cycle.

Creates a Ready-status backlog task from a pre-approved template, bypassing the full review cycle.

Schedules a non-coding work item — research inquiry, audit, survey, experiment, or written report — as a backlog entry. Single draft-and-critique cycle produces a phased execution checklist with shell-command gates. No compilation or test runner required.

Systematic technical debt quantification and management using SQALE methodology with value-effort prioritization, phased paydown roadmaps, and prevention strategies. Use when technical debt unmeasured or subjective, need objective prioritization, planning refactoring work, establishing debt prevention practices, or tracking debt trends over time. Provides 6 methodology components (measurement with SQALE index, categorization with code smell taxonomy, prioritization with value-effort matrix, phased paydown roadmap, trend tracking system, prevention guidelines), 3 patterns (SQALE-based quantification, code smell taxonomy mapping, value-effort prioritization), 3 principles (high-value low-effort first, SQALE provides objective baseline, complexity drives maintainability debt). Validated with 4.5x speedup vs manual approach, 85% transferability across languages (Go, Python, JavaScript, Java, Rust), SQALE industry-standard methodology.