amazon-working-backwards
Write a PR/FAQ for a product idea using Amazon's Working Backwards process. Start from the customer press release and work backward to what needs to be built. Use when validating whether an idea is worth building before committing to a spec.
69
62%
Does it follow best practices?
Impact
Pending
No eval scenarios have been run
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./product-skills/skills/amazon-working-backwards/SKILL.mdQuality
Discovery
89%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is a strong description that clearly identifies a specific methodology (Amazon's Working Backwards / PR/FAQ) and provides an explicit 'Use when' clause with a practical trigger scenario. The main weakness is that it could be more specific about the concrete deliverables or sub-actions involved in the process, such as drafting the press release, writing internal/external FAQs, or defining customer benefits.
Suggestions
Add more specific concrete actions like 'draft customer press release, write internal and external FAQ sections, define customer problem and benefits' to improve specificity.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | It names the domain (PR/FAQ, Amazon's Working Backwards process) and describes the core action (write a PR/FAQ, start from customer press release, work backward), but doesn't list multiple specific concrete sub-actions like 'draft press release, write FAQ section, define customer problem, outline metrics'. | 2 / 3 |
Completeness | Clearly answers both what ('Write a PR/FAQ for a product idea using Amazon's Working Backwards process, start from customer press release and work backward') and when ('Use when validating whether an idea is worth building before committing to a spec'). | 3 / 3 |
Trigger Term Quality | Includes strong natural trigger terms users would say: 'PR/FAQ', 'product idea', 'Working Backwards', 'press release', 'validating', 'worth building', 'spec'. These cover the key phrases someone exploring this process would use. | 3 / 3 |
Distinctiveness Conflict Risk | PR/FAQ and Amazon's Working Backwards process are highly specific and distinctive concepts unlikely to conflict with other skills. The combination of these terms creates a clear niche. | 3 / 3 |
Total | 11 / 12 Passed |
Implementation
35%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill provides a comprehensive Working Backwards PR/FAQ template but is significantly over-verbose, explaining concepts and structures that Claude already understands. The prompt template is a wall of instructional text that could be condensed to key structural requirements and a concrete example. The cross-references to other skills in the Tips section are a strength, but the lack of concrete output examples and explicit iteration loops weakens actionability and workflow clarity.
Suggestions
Cut the prompt template by 50%+ — remove explanations of what each section is for (e.g., 'This forces you to articulate the why now') and keep only the structural requirements and quality bars.
Add a concrete abbreviated example showing what a strong problem paragraph and a weak problem paragraph look like, so Claude can calibrate quality rather than relying on abstract instructions.
Add an explicit feedback loop: after the assessment, if any test fails, specify which sections to revise and re-evaluate before declaring the PR/FAQ complete.
Extract the full prompt template into a separate TEMPLATE.md file and keep SKILL.md as a concise overview with key principles and a reference to the template.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is extremely verbose. The entire prompt template is a massive block of instructions that Claude already knows how to execute given a brief description. Explaining what a press release is, what FAQs are, and how Working Backwards works is unnecessary context for Claude. The problem/solution paragraph instructions, quote formatting guidance, and reading level advice are all things Claude can infer from a much shorter prompt. This could be reduced by 60-70% without losing actionable content. | 1 / 3 |
Actionability | The skill provides a detailed template structure with specific sections and evaluation criteria, which is somewhat actionable. However, it's entirely prose-based instruction rather than concrete examples of good vs. bad output. There are no example PR/FAQs showing what a strong problem paragraph or customer quote actually looks like, which would make the guidance much more executable. | 2 / 3 |
Workflow Clarity | The workflow is implicitly sequential (write PR → write FAQs → do assessment → make kill/go decision), and the assessment section provides evaluation checkpoints. However, there's no explicit feedback loop — e.g., if the clarity test fails, what do you revise? If the kill test says rework, what's the iteration process? The Tips section hints at upstream/downstream connections but the core workflow lacks explicit validation-and-fix cycles. | 2 / 3 |
Progressive Disclosure | The Tips section references other skills (jobs-to-be-done, craft-discovery-synthesis, run-discovery, craft-spec, pre-mortem) which is good navigation. However, the massive prompt template is entirely inline when it could be split into a separate template file. There are no bundle files to support progressive disclosure, and the monolithic prompt block dominates the skill. | 2 / 3 |
Total | 7 / 12 Passed |
Validation
90%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 10 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 10 / 11 Passed | |
- Repository
- amplitude/builder-skills
- Commit
221ffaa
- Reviewed
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.