Turns work into multi-step assignments with nested substeps, fork delegation to long-running agent subprocesses, and inspectable session traces. Steps are defined from a catalog, assembled via presets, and driven through a restartable queue.

ace-demo records terminal sessions as reviewable proof-of-work artifacts, with sandbox-isolated YAML tapes and GitHub PR attachment for agent-driven evidence delivery.

Documentation management for teams and agents: track freshness from frontmatter, detect drift from git history, run LLM-powered analysis, and validate consistency across docs.

ace-git-secrets scans Git history for leaked credentials with gitleaks-backed detection, revokes supported tokens, rewrites compromised history, and blocks releases when secrets are still present.

Provides the dedicated ace-hitl package surface for human in the loop (HITL) semantics and CLI workflows.

Capture rough ideas quickly, store them as structured files, organize them with GTD-style folders, and manage them through a focused six-command CLI.

ace-lint provides Ruby-native linting for markdown (kramdown), YAML (Psych), Ruby files (StandardRB/RuboCop), and frontmatter without Node.js or Python dependencies.

Provides a single active prompt file with automatic history archiving. Write prompts in your editor, run ace-prompt-prep, get automatic archiving with optional context loading and LLM enhancement.

Create, review, and archive retrospectives from the command line with type-aware templates, task linkage, and health checks for long-running retrospective work.

Run focused, repeatable code reviews using configurable presets and multiple LLM models in parallel. Findings become trackable feedback items with a verify-apply-resolve lifecycle.

ace-sim executes preset-driven simulation chains across multiple providers so teams can validate ideas, review tasks, and compare synthesis outcomes before taking action.

Provides shared infrastructure for folder-based item management (tasks, ideas, etc.) across ace-* gems. Includes document loading, frontmatter parsing/serialization, filtering, sorting, directory scanning, and shortcut resolution.

Behavioral spec-based task management with compact IDs, subtask trees, AI-powered implementation plans, and automated health checks. Same CLI for developers and coding agents.

Wraps Minitest with smart grouping, cross-package resolution, and persistent reports so developers and coding agents can run focused checks, diagnose failures with context, and retain searchable execution history.

Runs package E2E scenarios through agents with reproducible setup, sandbox isolation, and structured reports for reliable validation workflows.