agentsclimarketplace

01 backlog

Skill andrefabricio/prodotypor/files/.claude/skills/01-backlog

An AI-assisted product development pipeline for Claude Code. Go from a product idea to Figma screens in 8 structured steps, with specialized AI agents handling product management, design research, and visual construction.

Install
npx -y skills add andrefabricio/prodotypor --skill 01-backlog

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Step 1 of 8 — Product backlog creation. Creates a feature backlog with epic grouping and user story titles from project context. Sub-commands: [continue] resume, [status] check progress, [update] force status refresh.

SKILL.md

4.2 KB, as published. Nobody here has run it

/01-backlog — Product Backlog Creation

Create or continue building the product backlog with epics, features, and user story titles.

Sub-command Router

Parse $ARGUMENTS:

  • Contains [status] → Jump to Status Report section
  • Contains [update] → Jump to Force Status Update section
  • Contains [continue] or empty → Jump to Execute section
  • Other text → Treat as context input (feature area, user need) and proceed to Execute

Prerequisites

Hard dependencies:

  1. Read docs/workflow-status.md — verify step 00 (Setup) status == completed
  2. Read CLAUDE.md — verify project context has non-placeholder content

If prerequisites fail:

  • If step 00 not completed: "Step 01 cannot start. Run /00-setup first to configure your project."
  • Do NOT proceed.

Validation:

  1. Read .impeccable.md — note brand personality (soft dependency, warn if missing)
  2. Check if docs/backlog/ directory exists (create if not)

Execute

Phase 1: Gather Context

  1. Read CLAUDE.md for project context, tech stack, and product scope
  2. Read .impeccable.md for brand personality and user personas
  3. If $ARGUMENTS contains specific context, note it as the focus area
  4. Check for existing backlog at docs/backlog/{{PROJECT_NAME}}-backlog.md — if present, this is a continuation

Phase 2: Delegate to Product Manager Agent

Spawn the product-manager agent with this prompt structure:

Create/update the product backlog for {{PROJECT_NAME}}.

Context: [project description from CLAUDE.md] User personas: [from .impeccable.md] Focus area: [from $ARGUMENTS if provided] Existing backlog: [content if continuing]

Output requirements:

  • Group features into epics (use templates/epic-template.md format)
  • Write user story titles for each feature (As a [persona], I want...)
  • Include phase grouping (MVP, Phase 1, Phase 2+)
  • Note dependencies between features
  • Save to docs/backlog/{{PROJECT_NAME}}-backlog.md

Phase 3: Validate Output

  1. Verify docs/backlog/{{PROJECT_NAME}}-backlog.md exists and is non-empty
  2. Check: has at least 1 epic grouping
  3. Check: has user story titles (not just feature names)
  4. If validation passes → proceed to Phase 4
  5. If validation fails → proceed to Error Recovery

Error Recovery

If validation fails:

  1. Identify the specific failure (which check failed, what's missing)
  2. Feed the error + partial output back to the agent:

    "The backlog did not meet requirements: [specific failure]. Current output: [summary of what exists]. Please fix: [specific instruction — e.g., 'add epic grouping' or 'convert feature names to user story titles']."

  3. Re-validate after the retry
  4. If retry also fails, mark step as blocked (reason) in docs/workflow-status.md
  5. Report the failure to the user with actionable guidance

Phase 4: Human Review Gate

  1. Update docs/workflow-status.md step 01 status to needs-review
  2. Report output summary to the user
  3. Prompt: "Review the backlog output. Run /01-backlog [continue] to approve and mark completed, or provide feedback for revisions."
  4. Only when the user runs [continue] after needs-review, transition status to completed

Phase 5: Report

Step 01 — Backlog Creation: [status]

Output: docs/backlog/{{PROJECT_NAME}}-backlog.md
Epics: [count]
Features: [count]
User stories: [count]

Next step: /02-prioritize

Status Report

  1. Read docs/workflow-status.md — extract step 01 section
  2. If backlog file exists, report: epic count, feature count, story count
  3. Report current status and any blockers

Force Status Update

  1. Re-scan docs/backlog/{{PROJECT_NAME}}-backlog.md
  2. Count epics, features, and stories
  3. Update docs/workflow-status.md step 01 with current state
  4. Report updated status

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.