Agent Watchtower Core v0.1.4

Quality patch: worker-run can record real result content, task-add supports in_progress tasks, --version is available, AGENTS structured instructions are added, release preflight runs in the PyPI workflow, and PyPI metadata is improved. Core public command count remains five.

Agent Watchtower Core v0.1.3

Documentation patch: clarify GitHub language navigation, remove time-limited trial wording, and align website version text. Core CLI behavior is unchanged.

v0.1.2

Documentation patch: clarify local demo wording, split English/Chinese presentation, and add platform support notes. No CLI behavior changes.

v0.1.1

Patch release: update public install docs now that PyPI publishing is live. Core CLI behavior is unchanged.

Agent Watchtower v0.1.0

Agent Watchtower v0.1.0

First alpha release of Agent Watchtower Core.

Agent Watchtower is a tiny local CLI for coding-agent continuity. It helps an interrupted coding agent resume by keeping a small local trail of:

• the current goal
• the next safe action
• the last run receipt
• the latest markdown artifact

Install from GitHub

python3 -m pip install "agent-watchtower-core @ git+https://github.com/un-n-smith/agent-watchtower.git@v0.1.0"

Or with pipx:

pipx install "git+https://github.com/un-n-smith/agent-watchtower.git@v0.1.0"

Quick demo

ROOT="$(mktemp -d)"
agent-watchtower --root "$ROOT" init
agent-watchtower --root "$ROOT" worker-status
agent-watchtower --root "$ROOT" worker-run
agent-watchtower --root "$ROOT" artifact-path

Public commands

• init
• task-add
• worker-status
• worker-run
• artifact-path

Boundaries

This release is a local continuity layer. It is not a daemon, scheduler, memory database, messaging platform, or full autonomous agent system.

Feedback is welcome. Interrupted-agent stories, install failures, and plain-language suggestions are especially useful.