qbl

Files and retrieves project backlog items with explicit ordering criteria and dependency tracking.

Updated Aug 14, 2026
One-click install
npx skills add https://github.com/QSchlegel/session-viz --skill qbl-qschlegel
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: qbl
Source: https://github.com/QSchlegel/session-viz/tree/main/plugins/session-viz/skills/qbl
Command: npx skills add https://github.com/QSchlegel/session-viz --skill qbl-qschlegel

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? Ad-hoc notes about future work get lost or resurface without context, and most backlog tools present an ordering they cannot justify. This Skill records what, where, and why for each item, then answers "what's next" while naming exactly which criteria separated the list and which separated nothing. ## Core Features & Use Cases - Bidirectional backlog: Push an item with its repo, branch, and an explicit --why, or pull an ordered list of what to pick up next. - Honest ordering: Ranks by recorded blockers, unblock counts, branch match, then age, and states outright when age alone determined the order. - Local and shared scopes: Keeps a per-project local file or pushes to a shared team queue, falling back to local with exit code 2 when the shared backend is unreachable. - Use Case: A developer finishes a session, files "the picker stub has drifted from the real stream" with a reason, and weeks later asks what to work on next, receiving a ranked list with its ordering basis printed. ## Quick Start Ask the assistant to run /qbl with a note about something to remember, or run /qbl with no argument to see what to work on next.

Frequently Asked Questions about qbl

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

FAQPage Schema
How do I save a task for later from the command line?▼

Run the qbl script with the task text as an argument and optionally add --why to record why it matters. The item is filed with the current repo, branch, and path automatically, so no manual context entry is needed.

How does the backlog decide what to work on next?▼

It applies four criteria in order: items with open blockers are held, then ranked by how many items they unblock, then by whether they were filed on the current branch, then oldest first. The output states which criteria actually separated items and which did not.

Can I share a backlog with my team?▼

Yes, the --shared flag pushes to a shared team queue instead of the local per-project file. It requires a collab-scoped token and an actor label; without them the command falls back to the local backlog and names the missing credential.

What happens when the shared backlog server is unreachable?▼

The command falls back to the local backlog, prints the connection failure reason, and exits with code 2 so scripts cannot mistake a partial answer for a complete one. A shared push that cannot land is refused entirely rather than silently filed locally.

What are the limitations of the shared backlog mode?▼

There is no deduplication, so pushing the same text twice files it twice. Also, closing shared tasks is not supported from this command because only accepted tasks can transition to done, which requires the separate team handoff tool.