subagent-driven-development
Use when executing implementation plans with independent tasks in the current session
48
36%
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 ./skills/subagent-driven-development/SKILL.mdQuality
Discovery
17%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 too vague and abstract to be useful for skill selection. It fails to describe any concrete actions or capabilities, uses no natural trigger terms a user would employ, and is so generic it could conflict with many other skills. The only positive aspect is the presence of a 'Use when' clause, though even that is poorly defined.
Suggestions
Add specific concrete actions the skill performs, e.g., 'Parallelizes independent subtasks from a plan, tracks completion status, and aggregates results' instead of the vague 'executing implementation plans'.
Include natural trigger terms users would actually say, such as 'parallel execution', 'run tasks simultaneously', 'batch processing', 'concurrent tasks', or whatever the actual use case is.
Clarify what 'implementation plans' and 'independent tasks' mean in practice—what types of tasks, what format of plans, and what 'current session' implies—to make the skill distinguishable from generic task execution skills.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description uses vague language like 'executing implementation plans' and 'independent tasks' without specifying any concrete actions. There are no specific capabilities listed—just abstract concepts. | 1 / 3 |
Completeness | It has a 'Use when' clause addressing when to use the skill, but the 'what does this do' part is extremely weak—'executing implementation plans with independent tasks' barely describes what the skill actually does. The 'when' is present but the 'what' is essentially missing. | 2 / 3 |
Trigger Term Quality | The terms 'implementation plans' and 'independent tasks' are not natural phrases users would say. Users are more likely to say things like 'run these tasks in parallel' or 'execute my plan'. The language is overly abstract and jargon-heavy. | 1 / 3 |
Distinctiveness Conflict Risk | 'Executing implementation plans' is extremely generic and could overlap with virtually any task execution, project management, or workflow skill. There are no distinct triggers that would differentiate this from other skills. | 1 / 3 |
Total | 5 / 12 Passed |
Implementation
54%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The skill has excellent workflow clarity with well-defined validation checkpoints, feedback loops, and error handling for subagent statuses. Progressive disclosure is well-handled with appropriate references to prompt templates and related skills. However, the content is significantly over-verbose — the Graphviz diagrams, lengthy advantages/comparison sections, and detailed example workflow consume many tokens while largely restating the same information. Actionability suffers from lack of concrete executable examples of how to actually dispatch subagents.
Suggestions
Cut the Graphviz diagrams and replace with a compact numbered list or brief ASCII flow — the textual process description and example workflow already convey the same information at lower token cost.
Remove or drastically condense the 'Advantages' section — Claude doesn't need to be sold on why to use the skill, it needs to know how.
Add a concrete, minimal example of the actual subagent dispatch mechanism (e.g., the tool call or command used to spawn a subagent with a prompt template).
Consolidate the 'Red Flags' section with the process description — many red flags are just negations of already-stated steps (e.g., 'don't skip reviews' restates 'do reviews').
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is extremely verbose at ~250+ lines. The Graphviz diagrams, while visually descriptive, are token-heavy and redundant with the textual process description. The 'Advantages' section explains benefits Claude doesn't need (e.g., 'Fresh context per task (no confusion)'). The 'When to Use' decision tree, example workflow, and red flags sections all contain significant redundancy with each other and with information Claude can infer. | 1 / 3 |
Actionability | The skill provides a clear process and references prompt templates (./implementer-prompt.md, etc.), but lacks executable code or commands. The example workflow is illustrative prose rather than copy-paste ready instructions. Key details like how to actually dispatch a subagent, what tool to use, or what the prompt templates contain are missing from this file. | 2 / 3 |
Workflow Clarity | The multi-step process is clearly sequenced with explicit validation checkpoints (spec compliance review → code quality review), feedback loops (reviewer finds issues → implementer fixes → re-review), and clear handling of edge cases (DONE, DONE_WITH_CONCERNS, NEEDS_CONTEXT, BLOCKED statuses). The ordering constraint (spec before quality) is explicitly called out. | 3 / 3 |
Progressive Disclosure | The skill appropriately references external prompt templates (implementer-prompt.md, spec-reviewer-prompt.md, code-quality-reviewer-prompt.md) and related skills (using-git-worktrees, writing-plans, etc.) with clear one-level-deep navigation. The main file serves as a coordination overview while delegating implementation details to referenced files. | 3 / 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
- obra/superpowers
- Commit
b7a8f76
- 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.