subagent-creator
Guide for creating AI subagents with isolated context for complex multi-step workflows. Use when users want to create a subagent, specialized agent, verifier, debugger, or orchestrator that requires isolated context and deep specialization. Works with any agent that supports subagent delegation. Triggers on "create subagent", "new agent", "specialized assistant", "create verifier". Do NOT use for Cursor-specific subagents (use cursor-subagent-creator instead).
86
82%
Does it follow best practices?
Impact
Pending
No eval scenarios have been run
Passed
No known issues
Quality
Discovery
100%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 skill description that covers all key dimensions well. It provides specific capabilities, natural trigger terms, explicit 'Use when' and 'Do NOT use' clauses, and clear disambiguation from a related skill. The description is concise yet comprehensive, making it easy for Claude to select appropriately from a large skill set.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description lists multiple concrete actions and concepts: creating AI subagents with isolated context, complex multi-step workflows, and specific agent types (verifier, debugger, orchestrator). It also includes a clear exclusion clause for Cursor-specific subagents. | 3 / 3 |
Completeness | Clearly answers both 'what' (guide for creating AI subagents with isolated context for complex multi-step workflows) and 'when' (explicit 'Use when' clause with specific scenarios, plus explicit 'Do NOT use' guidance for disambiguation). This is a thorough description. | 3 / 3 |
Trigger Term Quality | Excellent coverage of natural trigger terms: 'create subagent', 'new agent', 'specialized assistant', 'create verifier', 'debugger', 'orchestrator'. These are terms users would naturally use, and the explicit trigger list makes matching straightforward. | 3 / 3 |
Distinctiveness Conflict Risk | The description explicitly disambiguates from a related skill (cursor-subagent-creator) with a 'Do NOT use' clause, and the niche of general-purpose subagent creation with isolated context is well-defined and unlikely to conflict with other skills. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
64%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a solid, actionable skill with excellent concrete examples and templates that Claude can directly use. Its main weaknesses are verbosity (explaining concepts Claude already knows, four full inline patterns) and lack of a validation feedback loop in the creation workflow. Tightening the content and splitting patterns into a referenced file would significantly improve it.
Suggestions
Remove the 'What are Subagents?' section and the subagent-vs-skill decision tree—Claude already understands these concepts. Start directly with the structure/creation process.
Add an explicit validation step to the creation workflow, e.g., 'After creating the subagent, test it by delegating a sample task and verifying the output matches expectations. If not, refine the prompt and re-test.'
Move the four common patterns into a separate PATTERNS.md file and reference it from the main skill, keeping only one example inline as illustration.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill includes some unnecessary explanations (e.g., 'What are Subagents?' section explaining concepts Claude already knows, the decision tree for subagents vs skills). The common patterns section is useful but verbose—four full examples when two would suffice. The content could be tightened by ~30%. | 2 / 3 |
Actionability | Provides fully concrete, copy-paste-ready markdown templates with proper frontmatter, multiple complete subagent examples, specific metadata configuration guidance with good/bad description examples, and a clear output message format. The guidance is specific and executable. | 3 / 3 |
Workflow Clarity | The creation process has clear numbered steps (Define Purpose → Configure Metadata → Write Prompt), and there's a quality checklist at the end. However, there's no explicit validation/verification step in the creation workflow itself—no 'test the subagent' step or feedback loop for checking if the created subagent works correctly before finalizing. | 2 / 3 |
Progressive Disclosure | The content is well-structured with clear headers and sections, but it's monolithic—all four common patterns are fully inline rather than referenced from separate files. For a skill this long (~200+ lines), the patterns and detailed examples could be split into a referenced PATTERNS.md file, keeping the main skill leaner. | 2 / 3 |
Total | 9 / 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.
- Repository
- tech-leads-club/agent-skills
- Commit
81e7e0d
- 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.