framework-health

Monitors framework state via file existence checks and backlog signals without blocking progress.

Updated Apr 13, 2026
One-click install
npx skills add https://github.com/scoots31/engineering-playbook --skill framework-health-scoots31
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: framework-health
Source: https://github.com/scoots31/engineering-playbook/tree/main/skills/framework-health
Command: npx skills add https://github.com/scoots31/engineering-playbook --skill framework-health-scoots31

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Long-running, multi-phase software projects lose integrity silently — missing gate outputs, stale handoffs, outdated backlogs, and skipped review steps go unnoticed until they block work. This Skill runs as a background health monitor that detects these gaps early and surfaces them with a specific recovery path, without ever halting progress. ## Core Features & Use Cases - Signal-based health checks: Verifies phase completion through file existence checks and the backlog At a Glance section instead of reading full documents, keeping token usage low. - Lifecycle activation points: Runs at mode activation (framework version check, project state read, 21-day long-gap detection), between phases (gate integrity), during sessions (milestone lines, one hygiene reminder), and at session end (handoff currency, backlog accuracy, open issues). - Non-blocking issue surfacing: Reports one issue at a time with a concrete recovery path, logs deferred items to the session handoff, and stays silent when everything is healthy. - Use Case: When moving from design sprint to design review, the monitor detects that the deferred decisions log is missing and asks whether to produce it before the next phase starts — the user can proceed, defer, or fix it immediately. ## Quick Start Activate guided or piloted mode and let the health monitor check the handoff, backlog, and expected phase outputs before starting new work.

Frequently Asked Questions about framework-health

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I monitor a multi-phase AI coding workflow for missing outputs?

Use signal-based health checks that verify expected files exist for each completed phase rather than reading full documents. Check the handoff file and backlog summary counts, then surface one gap at a time with a specific recovery path.

What is the best way to detect stale project state between sessions?

Compare the date in the session handoff file to today at session start. Gaps of 21 or more days trigger a re-entry check covering codebase state, environment, and open decisions before new work begins.

Does the health monitor block work when it finds a problem?

No, it never blocks progress. It surfaces one issue with a recovery path, and if the user defers it, the item is logged to the session handoff and checked again at session end rather than repeated mid-session.

How do I check for framework updates without interrupting a session?

Run a git fetch and rev-list count against origin/main once at mode activation, before reading project documents. If updates exist, ask once whether to pull; never repeat the check mid-session.

When should a phase test readiness reminder fire?

Only when the backlog summary shows all current-phase slices at Done with none In Progress, In QA, or In Review, and no open triage items remain. It fires once and does not repeat if the user is not ready.