cursor-agent-supervisor
Offloading tasks with a well-defined scope to sub-agents, for instance to use a sub-agent to implement a set of specs. Use this skill whenever a task should not need a broad knowledge of the whole project
76
66%
Does it follow best practices?
Impact
91%
1.59xAverage score across 3 eval scenarios
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./cursor-agent-supervisor/SKILL.mdQuality
Discovery
40%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
The description conveys the general concept of delegating work to sub-agents but lacks specificity in both capabilities and trigger conditions. The 'when' guidance is defined by a vague negative condition rather than concrete scenarios, making it difficult for Claude to reliably select this skill. The description would benefit from concrete examples of task types and more natural trigger terms.
Suggestions
Add specific concrete actions the skill performs, e.g., 'Spawns sub-agents to implement features from specs, run isolated test suites, refactor individual modules, or perform file-scoped changes.'
Replace the vague 'when' clause with explicit trigger scenarios, e.g., 'Use when the user asks to delegate, parallelize, or break down implementation tasks into independent units that can be handled in isolation.'
Include natural trigger terms users would say, such as 'delegate', 'subtask', 'parallel', 'spawn agent', 'break down', 'independent task'.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description uses vague language like 'offloading tasks' and 'well-defined scope' without listing concrete actions. The example of 'implement a set of specs' is a single, loosely defined example rather than specific capabilities. | 1 / 3 |
Completeness | It addresses both 'what' (offloading tasks to sub-agents) and 'when' (tasks that don't need broad project knowledge), but the 'when' clause is vague and defined by negation rather than explicit positive triggers. The guidance is too abstract to reliably match user requests. | 2 / 3 |
Trigger Term Quality | The term 'sub-agent' is a relevant keyword, and 'specs' provides some trigger value. However, it misses common natural variations users might say like 'delegate', 'parallel tasks', 'spawn agent', 'break down work', or 'subtask'. | 2 / 3 |
Distinctiveness Conflict Risk | The concept of 'sub-agents' provides some distinctiveness, but 'offloading tasks with well-defined scope' is broad enough to potentially overlap with any task decomposition or delegation skill. The trigger condition ('should not need broad knowledge') is subjective and hard to distinguish. | 2 / 3 |
Total | 7 / 12 Passed |
Implementation
92%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a high-quality skill that provides clear, actionable guidance for delegating tasks to sub-agents. It excels at conciseness (no unnecessary explanations), actionability (concrete commands and templates), and workflow clarity (clear sequence with validation and error recovery). The only minor weakness is that the jj-todo-workflow reference is not linked, and the content could benefit slightly from cross-references to related skills.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The content is lean and efficient. Every section serves a purpose—no explanation of what sub-agents are conceptually, no padding about why delegation is useful. It assumes Claude understands the concepts and focuses on the specific tool invocations and workflow patterns. | 3 / 3 |
Actionability | Provides concrete, copy-paste ready commands (cursor-agent invocations, timeout settings), a specific task description template with placeholders, and explicit do/don't lists. The model selection tip with the error-listing trick is a practical, actionable detail. | 3 / 3 |
Workflow Clarity | The workflow is clearly sequenced: prepare the environment → craft instructions → invoke sub-agent with timeout → verify results. The 'After Sub-agent Completes' section provides explicit validation steps (check diff, review code, run tests), and 'When Things Go Wrong' provides error recovery feedback loops. | 3 / 3 |
Progressive Disclosure | The content is well-structured with clear sections and headers, but everything is inline in a single file. For a skill of this length (~80 lines of content), this is mostly fine, but the task description template and troubleshooting sections could potentially be split out. The reference to 'jj-todo-workflow' is mentioned but not linked. | 2 / 3 |
Total | 11 / 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
- YPares/agent-skills
- Commit
aed1afb
- 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.