Release Notes & Changelog

Relevant source files

The following files were used as context for generating this wiki page:

• CHANGELOG.md
• README.md
• src/main.ts
• src/queue-controller.ts
• package.json
• SECURITY.md

Introduction

PlanQueue is an OMP (Oh My Pi) extension designed to streamline the interaction between developers and coding agents by transforming session notes into a FIFO prompt queue. The "Release Notes & Changelog" serves as the authoritative record of the project's evolution, documenting features, improvements, and breaking changes. It ensures transparency for users and maintainers regarding the transition from manual prompt management to an automated, persistent queue system.

The project adheres to Semantic Versioning and follows the Keep a Changelog format. Versioning began with the first public release, marking the transition from internal iterations to a stable public identity.

Sources: CHANGELOG.md:1-10, README.md:1-15, package.json:1-10

---

Version 0.1.0 (First Public Release)

The initial public release, version 0.1.0, introduced the core architecture of PlanQueue, establishing it as a persistent notes panel that functions as an agent-driven task queue.

Core Features and User Interface

The release featured a session-notes panel rendered below the status line. This panel is not just for storage; it acts as a functional FIFO queue where Markdown checkboxes represent task states.

Glyph	Markdown State	Task Meaning
☐	- [ ]	Pending
▸	- [>]	In-flight
✓	- [x]	Done

Sources: CHANGELOG.md:12-25, README.md:50-60, src/main.ts:145-155

Automation and Human-in-the-Loop (HITL)

The release introduced two primary modes of operation: manual stepping and auto-run.

• Manual Step: Triggered via Ctrl+↓, it dispatches the next pending task.
• Auto-run: Triggered via Ctrl+Shift+↓, it automatically feeds tasks to the agent upon every turn settlement.
• Barriers: The use of --- lines acts as a checkpoint, pausing the queue (especially during auto-run) to allow for human review.

The following diagram illustrates the flow of a task from the queue to the agent:

LLM Tools Integration

To allow the agent to manage its own queue, two LLM tools were introduced:

1. note_add: Appends a single task to the note.
2. make_note: Decomposes a high-level goal into a multi-step prompt queue with optional barriers.

Sources: CHANGELOG.md:37-41, src/main.ts:200-300

---

Technical Architecture Evolution

The transition to public release involved migrating from legacy storage systems to a structured directory hierarchy under ~/.planqueue/.

Storage and Persistence Hierarchy

Notes are organized by repository, branch, and session ID to ensure context is never mixed across different work environments.

Security and Runtime Scope

The project established a strict security policy starting with 0.1.0. The extension is confined to local I/O, specifically within the ~/.planqueue/ tree, and performs no network telemetry. It utilizes OSC 52 escape sequences for clipboard actions, ensuring compatibility across SSH sessions.

Sources: SECURITY.md:20-45, src/main.ts:40-55

---

Configuration and Shortcuts

Version 0.1.0 introduced a global configuration file at ~/.planqueue/config.json. This allows users to override default keybindings, which is particularly useful for terminal environments that may not support specific key combinations (e.g., Ctrl+Shift+↓).

Configuration Key	Default Shortcut	Description
editNotes	ctrl+n	Opens the multi-line note editor
queueStep	ctrl+down	Dispatches next task or passes barrier
queueToggleAuto	ctrl+shift+down	Toggles the auto-run mode

Sources: CHANGELOG.md:42-43, README.md:110-125, src/main.ts:175-190

Summary

The release notes for PlanQueue document the transition of the project into a public OMP extension. From the initial release of version 0.1.0, the focus has been on providing a robust, persistent FIFO queue for AI interactions, supported by a structured storage system, integrated LLM tools, and flexible configuration. This foundation ensures that developers can "write the plan once" and maintain high-velocity workflows with coding agents.

Sources: CHANGELOG.md:50-52, README.md:5-10