project-development
This skill should be used when the user asks to "start an LLM project", "design batch pipeline", "evaluate task-model fit", "structure agent project", or mentions pipeline architecture, agent-assisted development, cost estimation, or choosing between LLM and traditional approaches.
52
39%
Does it follow best practices?
Impact
Pending
No eval scenarios have been run
Advisory
Suggest reviewing before use
Optimize this skill with Tessl
npx tessl skill review --optimize ./skills/project-development/SKILL.mdQuality
Discovery
37%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 essentially a 'when to use' clause with no 'what it does' clause. While the trigger terms are strong and varied, the complete absence of concrete capability descriptions makes it impossible for Claude to understand what this skill actually produces or accomplishes. The description needs a clear statement of actions and outputs before the trigger guidance.
Suggestions
Add a leading sentence describing concrete actions, e.g., 'Scaffolds LLM-based projects, designs batch and streaming pipelines, estimates API costs, evaluates task-model fit, and structures agent architectures.'
Restructure to follow the pattern: '[What it does]. Use when [triggers].' to ensure both halves are clearly present.
Specify outputs or deliverables (e.g., 'produces architecture diagrams, cost breakdowns, and project plans') to further distinguish from other LLM-related skills.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description lists no concrete actions or capabilities—it only describes when to use the skill via trigger phrases, but never explains what the skill actually does (e.g., 'generates project scaffolding', 'produces cost estimates', etc.). | 1 / 3 |
Completeness | The description answers 'when' extensively but completely fails to answer 'what does this do'—there is no explanation of the skill's capabilities, outputs, or actions. Both halves need to be present for a higher score. | 1 / 3 |
Trigger Term Quality | The description includes a rich set of natural trigger terms users would say: 'start an LLM project', 'design batch pipeline', 'evaluate task-model fit', 'structure agent project', 'pipeline architecture', 'agent-assisted development', 'cost estimation', 'choosing between LLM and traditional approaches'. These cover multiple natural phrasings. | 3 / 3 |
Distinctiveness Conflict Risk | The trigger terms are fairly specific to LLM project planning and pipeline design, which helps distinguish it, but without stating what the skill actually does, it could overlap with other LLM-related skills (e.g., an LLM coding skill, an LLM evaluation skill). | 2 / 3 |
Total | 7 / 12 Passed |
Implementation
42%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill covers a broad methodology topic comprehensively but suffers from significant verbosity—it explains many concepts Claude already understands (LLM strengths/weaknesses, idempotency, caching benefits) and provides rationale for nearly every statement. The progressive disclosure and organization are strong, with clear references to related skills and external files. However, the lack of executable code examples and the excessive explanatory text significantly reduce its effectiveness as a skill file.
Suggestions
Cut explanatory rationale clauses (e.g., 'because X') by at least 50%—Claude understands why idempotency and caching are useful. Remove the task-model fit tables entirely or reduce to a brief checklist, as Claude already knows LLM strengths and weaknesses.
Add executable code snippets for key patterns: a concrete Python example of the file-system state machine check, a working cost estimation calculation, and a minimal pipeline stage implementation.
Add explicit validation checkpoints between pipeline stages (e.g., 'Verify parsed.json has expected schema before proceeding to render') with concrete checks rather than abstract guidance.
Reduce the Gotchas section to 3-4 items that are genuinely non-obvious; items like 'estimate costs early' and 'don't skip validation' are standard engineering advice that doesn't need elaboration.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is extremely verbose at ~300+ lines, explaining many concepts Claude already knows (what batch processing is, what idempotency means, general software architecture principles). Tables explaining why LLMs are good at synthesis or bad at math are common knowledge for Claude. Phrases like 'because separation of deterministic and non-deterministic stages enables fast iteration' are explanatory padding. | 1 / 3 |
Actionability | The skill provides structured guidance and a project planning template, but lacks executable code examples. The pipeline structure is described abstractly ('acquire -> prepare -> process -> parse -> render') without concrete implementation. The file system state pattern shows directory structure but no actual code to implement it. Cost formula is given but no concrete example with real numbers beyond the case study summary. | 2 / 3 |
Workflow Clarity | The Project Planning Template provides a clear 5-step sequence, and the pipeline stages are well-ordered. However, validation checkpoints are mostly implicit—there's no explicit 'validate before proceeding' step with concrete checks between pipeline stages. The manual prototype step is described but lacks a clear feedback loop for what to do when it fails beyond 'stop'. | 2 / 3 |
Progressive Disclosure | The skill effectively references external files (case-studies.md, pipeline-patterns.md) and related skills (tool-design, multi-agent-patterns, evaluation) with clear 'Read when' guidance. Content is organized into logical sections with a clear hierarchy from core concepts to detailed topics to practical guidance. References are one level deep and well-signaled. | 3 / 3 |
Total | 8 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
- Commit
7a95d94
- 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.