Backlog Prioritizer

Prioritize your backlog with RICE, MoSCoW, or Kano — framework selected based on your context.

Install
cmdop skills install agensi-backlog-prioritizer

Backlog Prioritizer is a Claude Code skill that applies structured prioritization frameworks to a product or engineering backlog. It is delivered as a SKILL.md file installed in a Claude Code project. Once installed, it activates on natural-language triggers such as “prioritize my backlog”, “apply RICE”, “help me decide what to build next”, or “rank these features”.

The skill begins by reading the stated goal, target audience, and constraints before recommending one of three frameworks. RICE is used for objective, numerically scored prioritization; MoSCoW is applied when the focus is release scoping; Kano is selected for discovery contexts where understanding user delight versus dissatisfaction matters. The skill auto-calculates RICE scores and assigns MoSCoW and Kano categories, each accompanied by a written rationale.

Beyond scoring, it flags dependencies between items, highlights effort outliers, and marks any item whose score rests on assumptions rather than confirmed data. It also surfaces conflicts between the scored ranking and a team’s stated instincts without overriding judgment — keeping the output advisory rather than prescriptive.

The final output is a ranked, decision-ready table. The skill does not connect to any external data source, project management tool, or ticketing system; all input is provided through Claude Code conversation.

Use cases

  • Score a feature list using RICE when preparing a quarterly roadmap
  • Apply MoSCoW categories to scope a release under a fixed deadline
  • Run a Kano analysis during early discovery to separate must-haves from delighters
  • Flag backlog items that depend on unfinished work before committing to a sprint
  • Identify items scored on assumptions so they can be validated before prioritization is finalised
  • Get a ranked table from an unordered list of features for a stakeholder review

When to use it

  • When working inside Claude Code and needing a structured way to rank a backlog
  • When the right prioritization framework is unclear and context-based recommendation is needed
  • When a decision-ready ranked table with written rationale is required for stakeholder alignment
  • When dependencies or effort outliers in a backlog need to be surfaced explicitly

When not to use it

  • When the project does not use Claude Code, as the skill requires SKILL.md installation in a Claude Code project
  • When backlog data lives in an external tool such as Jira or Linear and automated sync is needed — the skill has no external integrations
  • When a fully automated, code-executed scoring pipeline is required rather than a conversational output
  • When a license-verified or enterprise-audited package is required — no license information is provided for this skill