recipe-fullstack-build
Execute decomposed fullstack tasks with layer-aware agent routing
26
17%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./skills/recipe-fullstack-build/SKILL.mdQuality
Discovery
0%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 description is heavily laden with technical jargon and buzzwords ('layer-aware agent routing', 'decomposed fullstack tasks') without conveying any concrete actions or use cases. It fails to answer what the skill specifically does, when it should be used, and uses language no user would naturally employ to trigger it.
Suggestions
Replace abstract jargon with concrete actions, e.g., 'Generates frontend components, API endpoints, and database schemas for fullstack features' instead of 'decomposed fullstack tasks with layer-aware agent routing'.
Add an explicit 'Use when...' clause with natural trigger terms, e.g., 'Use when the user asks to build a fullstack feature, create an API with a frontend, or scaffold a web application across multiple layers.'
Include specific file types, frameworks, or technologies this skill handles to distinguish it from other development skills and improve trigger term coverage.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description uses vague, abstract language like 'decomposed fullstack tasks' and 'layer-aware agent routing' without listing any concrete actions. No specific capabilities are described. | 1 / 3 |
Completeness | The description vaguely hints at 'what' but uses abstract language that doesn't clearly explain what the skill does. There is no 'when' clause or explicit trigger guidance whatsoever. | 1 / 3 |
Trigger Term Quality | The terms used are technical jargon ('layer-aware agent routing', 'decomposed') that users would never naturally say. There are no natural keywords a user would use when requesting this skill. | 1 / 3 |
Distinctiveness Conflict Risk | 'Fullstack tasks' is extremely broad and could overlap with virtually any development-related skill. The jargon-heavy phrasing doesn't carve out a clear niche. | 1 / 3 |
Total | 4 / 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 is a complex orchestration skill that defines a thorough multi-phase workflow with good escalation handling and verification steps. Its main weaknesses are excessive verbosity (the Consumed Task Set filtering logic is repeated three times nearly verbatim), a mislabeled '4-step' cycle that actually has 6 steps, and a monolithic structure that would benefit from extracting repeated logic into referenced files. The actionability is moderate — it specifies what to do but relies heavily on external references without providing complete executable examples.
Suggestions
Extract the Consumed Task Set computation logic into a single referenced section or file and refer to it by name instead of repeating the full filtering rules three times.
Fix the '4-Step Cycle' label to match the actual 6 steps, or restructure to truly be 4 steps (e.g., combine steps 4-5 into one quality-check step).
Add a concrete, complete example of a subagent invocation (Agent tool call with all parameters filled in) for at least one routing case to improve actionability.
Move the detailed file-pattern exclusion rules and Implementation Readiness Check logic into a reference file to reduce the main skill's token footprint.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is extremely verbose at ~200+ lines with extensive procedural detail that could be significantly compressed. Many sections repeat the same filtering logic (Consumed Task Set pattern is restated multiple times), and the Implementation Readiness Check table, Task Generation Decision Flow table, and Pre-execution Prerequisites could all be dramatically tightened. Claude doesn't need this level of hand-holding for file pattern matching or conditional branching. | 1 / 3 |
Actionability | The skill provides concrete agent routing tables, specific file patterns, and structured response field names to check (e.g., `status: "escalation_needed"`, `readyForQualityCheck: true`). However, it lacks executable code examples — the subagent invocations are described but not shown as complete tool calls, and the git/filesystem operations are mentioned without exact commands. The task cycle references step numbers inconsistently (describes 4-step but lists 6 steps). | 2 / 3 |
Workflow Clarity | The multi-step workflow is clearly sequenced with explicit phases (Pre-execution → Task Decomposition → Execution Cycle → Post-Implementation → Cleanup → Report), and includes validation checkpoints and escalation paths. However, the 4-step cycle is labeled as '4-Step' but actually contains 6 steps, which is confusing. The feedback loops for quality-fixer and code-verifier are well-defined, but the step numbering inconsistency and the repeated Consumed Task Set computation logic across sections reduce clarity. | 2 / 3 |
Progressive Disclosure | The skill references external files (`references/monorepo-flow.md`, `subagents-orchestration-guide`) appropriately, and the references are one level deep. However, no bundle files are provided, making it impossible to verify these references resolve correctly. The skill itself is monolithic — much of the Consumed Task Set filtering logic and the Implementation Readiness Check could be extracted to a reference file, keeping the main skill leaner. | 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
- shinpr/claude-code-workflows
- Commit
adf2e4d
- 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.