cursor-subagent-creator
Creates Cursor-specific AI subagents with isolated context for complex multi-step workflows. Use when creating subagents for Cursor editor specifically, following Cursor's patterns and directories (.cursor/agents/). Triggers on "cursor subagent", "cursor agent". Do NOT use for generic subagent creation outside Cursor (use subagent-creator instead).
76
71%
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 ./packages/skills-catalog/skills/(creation)/cursor-subagent-creator/SKILL.mdQuality
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 an excellent skill description that clearly defines its scope, provides explicit trigger terms, and proactively disambiguates from a related skill. The 'Do NOT use' clause with a redirect to the alternative skill is a particularly strong pattern for avoiding conflicts in a multi-skill environment.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists concrete actions: creates subagents with isolated context for complex multi-step workflows, follows Cursor's patterns and directories (.cursor/agents/). Also specifies what NOT to use it for, adding clarity. | 3 / 3 |
Completeness | Clearly answers both 'what' (creates Cursor-specific AI subagents with isolated context for complex multi-step workflows) and 'when' (explicit 'Use when' clause with triggers, plus a 'Do NOT use' clause for disambiguation). Excellent completeness. | 3 / 3 |
Trigger Term Quality | Includes natural trigger terms users would say: 'cursor subagent', 'cursor agent', 'Cursor editor', 'subagent'. Also mentions the specific directory pattern (.cursor/agents/) which helps with file-based triggers. Explicitly distinguishes from 'generic subagent creation'. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive — explicitly scoped to Cursor editor only, references specific directory (.cursor/agents/), and includes a 'Do NOT use' clause pointing to the alternative skill (subagent-creator) for generic cases. This virtually eliminates conflict risk. | 3 / 3 |
Total | 12 / 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.
The skill is highly actionable with excellent concrete templates and examples, but is severely bloated—likely 3-4x longer than necessary. It suffers from significant redundancy (anti-patterns repeated, concepts over-explained) and dumps everything into a single monolithic file rather than using progressive disclosure. The workflow is reasonably clear but lacks explicit verification steps for the creation process itself.
Suggestions
Cut content by 60%+: Remove the 'What are Subagents?' section, the decision tree, performance trade-offs, and redundant anti-patterns. Keep only 1-2 example subagents inline and move the rest to a separate EXAMPLES.md file.
Split into progressive disclosure: Create separate files like `PATTERNS.md` (for the 6 common patterns), `REFERENCE.md` (for field configuration details), and keep SKILL.md as a concise overview with the quick template and creation process.
Add a verification step to the creation workflow: After step 5 (write the prompt), add an explicit step to test invocation and confirm the Agent delegates correctly, rather than burying 'test the description' in best practices.
Remove explanations of concepts Claude already knows: Drop explanations of kebab-case, what isolated context means, what parallel execution is, and the Skills vs Subagents vs Commands decision tree.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Extremely verbose at ~500+ lines. Explains concepts Claude already knows (what subagents are, what isolated context means, basic concepts like kebab-case). Massive redundancy: anti-patterns section repeats the AVOID section nearly verbatim. The 'What are Subagents?' section, decision tree explanation, and performance trade-offs table all explain things Claude can infer. Six full example subagents when 1-2 would suffice. | 1 / 3 |
Actionability | Provides fully concrete, copy-paste ready templates and examples. The file format, frontmatter fields, directory locations, invocation syntax, and complete example subagents are all specific and executable. The quick template and complete examples give Claude everything needed to create a subagent. | 3 / 3 |
Workflow Clarity | The 5-step creation process is clearly sequenced and the quality checklist provides validation. However, there are no explicit validation/verification checkpoints within the creation workflow itself—no step to test that the created subagent actually works or that the description triggers delegation correctly. The 'Test the description' item is buried in best practices rather than integrated into the workflow. | 2 / 3 |
Progressive Disclosure | Monolithic wall of text with no references to external files. The six complete subagent examples, detailed field configuration, performance considerations, and complete examples could all be split into separate reference files. Everything is inline in one massive document with no layered navigation structure. | 1 / 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 |
|---|---|---|
skill_md_line_count | SKILL.md is long (726 lines); consider splitting into references/ and linking | Warning |
Total | 10 / 11 Passed | |
- Repository
- tech-leads-club/agent-skills
- Commit
906a57d
- 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.