Status: Active Owner: Maintainers Source of truth: this document for documentation navigation and page ownership Parent: README
CueLoop is a Rust CLI plus SwiftUI macOS app for queue-driven, auditable AI coding agent work. Use this page as the map; each linked page owns its own detailed contract.
| If you are... | Start with | Then read |
|---|---|---|
| Evaluating CueLoop quickly | README | Evaluator Path, Local Smoke Test |
| Installing CueLoop in a repo | Quick Start | Configuration, CLI Reference |
| Learning the task ledger | Agent Usage Guide | Queue, Task System, Task Schema and Field Reference |
| Learning runner-backed workflow | Architecture Overview | Phases, Runners, Supervision |
| Running CueLoop day to day | CLI Reference | Troubleshooting, Feature Guides |
| Operating or releasing this project | CI and Test Strategy | Project Operating Constitution, Release Runbook |
| Building app or automation integrations | Machine Contract | App Feature Guide, Pi Integration |
| Using CueLoop as an already-running agent | Agent Usage Guide | Machine Contract, Queue, Task Schema and Field Reference |
These pages are the active sources of truth. Legacy URLs remain as navigation bridges when useful, but they should not accumulate new reference material.
| Topic | Canonical page |
|---|---|
| Product overview and fastest value proof | README |
| Install, init, and first local checks | Quick Start |
| Commands and flags | CLI Reference |
| Configuration, precedence, trust, profiles, plugins, and integrations | Configuration |
| Runtime architecture and trust boundaries | Architecture Overview |
| Queue file operations, ordering, locking, repair, archive, import/export | Queue |
| Task docs index | Task System |
| Task JSON fields and schema examples | Task Schema and Field Reference |
| Status transitions and priority semantics | Task Lifecycle and Priority |
| Dependencies, blocking, relations, duplicates, and hierarchy | Task Relationships |
| Creating, editing, templating, cloning, importing, and batching tasks | Task Operations |
| Multi-phase execution | Phases |
| Runner orchestration | Runners |
| CI gates, git oversight, and human-in-the-loop supervision | Supervision |
| Prompt overrides | Prompts |
| Session recovery | Session Management |
| Parallel execution | Parallel |
| Background automation | Daemon and Watch |
| Webhooks and notifications | Webhooks, Notifications |
| Security model | Security Model, Security Features |
Use these when you want a guided human-readable path rather than a machine contract or maintainer runbook.
- Quick Start: shortest install/init/inspect path.
- Getting Started: guided orientation with links to deeper docs.
- Evaluator Path: reviewer-friendly route through proof and validation.
- Local Smoke Test: deterministic no-runner validation.
- CueLoop Dogfood Harness: repeatable end-to-end fixture project.
- Advanced Usage Guide: power-user workflows, profiles, plugins, automation, and optimization.
The Feature Guides page is the feature-area index. Common entry points:
- CLI Reference
- Configuration
- Machine Contract
- Environment Variables
- Error Handling Guidelines
- Support Policy
- Versioning Policy
- Decisions
- Roadmap Archive
- Project Operating Constitution
- CI and Test Strategy
- Public Readiness Checklist
- Release Runbook
- Full Release Guide
These are for coding agents or app/automation clients, not first-time human onboarding.
- Agent Usage Guide
- Machine Contract
- Project Operating Constitution
- Configuration Trust and Precedence
Archive and audit docs are useful for history, review evidence, and follow-up planning. Most are point-in-time artifacts; active behavior is defined by the canonical docs above and generated schemas. The current stack audit remains an active baseline document until a newer stack audit supersedes it.
- Archive and audit policy
- Stack Audit (2026-04): current toolchain/dependency baseline
- Thermo-Nuclear Code Quality Review (2026-05-21): point-in-time maintainability review
- Comprehensive Codebase Audit (2026-03-31): point-in-time codebase audit
- CueLoopMac Settings Window Investigation (2026-03-13): resolved investigation notes
- Stack Audit (2026-03): older baseline kept for comparison
- Queue:
.cueloop/queue.jsonc - Done archive:
.cueloop/done.jsonc - Project config:
.cueloop/config.jsonc - Prompt overrides:
.cueloop/prompts/ - Runtime migration: use
cueloop migrate runtime-dir --checkbefore applying supported old-state migrations.
GNU Make >= 4 is required for project targets. On macOS, install with
brew install makeand usegmakeunless your PATH already exposes GNU Make asmake.
Use docs/guides/ci-strategy.md as the canonical validation guide.
Routine branch gate:
make agent-ciFinal ship/release gate:
make release-gateLower-level targets such as ci-docs, ci-fast, ci, and macos-ci still exist, but most contributors should treat them as internal tiers behind make agent-ci rather than commands to choose among day to day.
Routing uses only the current uncommitted working tree (including untracked paths); commits already on the branch do not change the tier. To debug routing, run scripts/agent-ci-surface.sh --target and --reason from the repo root. Changes to scripts/agent-ci-surface.sh or path allowlists in scripts/lib/release_policy.sh should stay aligned with contract coverage in crates/cueloop/tests/agent_ci_surface_contract_test.rs (see docs/guides/ci-strategy.md).