recipe-build
Execute decomposed tasks in autonomous execution mode
35
31%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./skills/recipe-build/SKILL.mdQuality
Discovery
0%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 extremely vague and fails on all dimensions. It uses abstract jargon ('decomposed tasks', 'autonomous execution mode') without specifying what kinds of tasks, what domain it operates in, or when Claude should select it. It would be nearly impossible for Claude to reliably choose this skill from a pool of alternatives.
Suggestions
Replace abstract language with concrete actions: specify what types of tasks are executed (e.g., 'Runs multi-step build pipelines, test suites, and deployment scripts').
Add an explicit 'Use when...' clause with natural trigger terms that users would actually say (e.g., 'Use when the user asks to run a sequence of steps, automate a workflow, or execute a multi-part plan').
Define the specific domain or niche this skill covers to distinguish it from other skills that also perform tasks (e.g., CI/CD, data processing, code generation).
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description uses vague, abstract language like 'decomposed tasks' and 'autonomous execution mode' without specifying any concrete actions. There are no specific capabilities listed. | 1 / 3 |
Completeness | The description weakly addresses 'what' with abstract language and completely lacks any 'when' clause or explicit trigger guidance. | 1 / 3 |
Trigger Term Quality | The terms 'decomposed tasks' and 'autonomous execution mode' are technical jargon that users would not naturally say. No common, natural keywords are present. | 1 / 3 |
Distinctiveness Conflict Risk | 'Execute decomposed tasks' is extremely generic and could apply to virtually any skill that performs multi-step work, creating high conflict risk with many other skills. | 1 / 3 |
Total | 4 / 12 Passed |
Implementation
62%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-structured orchestration skill with excellent workflow clarity, featuring explicit decision tables, validation checkpoints, and error recovery loops. Its main weaknesses are moderate verbosity from repeated filtering rules (the Consumed Task Set pattern is restated multiple times) and reliance on external subagent references without concrete invocation examples. The skill would benefit from deduplicating the task-file filtering logic and providing at least one complete Agent tool invocation example.
Suggestions
Deduplicate the Consumed Task Set filtering rules by defining them once and referencing that definition, rather than restating the exclusion patterns in multiple sections.
Add a concrete, complete example of an Agent tool invocation with actual parameters and a sample response showing how to parse the status fields (e.g., a full task-executor call and response).
Consider extracting the Pre-execution Prerequisites section (Implementation Readiness Check, Consumed Task Set computation, Task Generation Decision Flow) into a separate referenced file to improve progressive disclosure.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is quite lengthy and includes some redundant explanations (e.g., the Consumed Task Set computation is described multiple times with near-identical filtering rules). However, most content is necessary given the complexity of the orchestration workflow, and it avoids explaining basic concepts Claude already knows. | 2 / 3 |
Actionability | The skill provides structured steps, decision tables, and specific subagent invocation parameters, which is good. However, it relies heavily on references to other skills/subagents (task-decomposer, quality-fixer, code-verifier, etc.) without concrete executable examples of the actual Agent tool invocations or response parsing. The guidance is specific but not fully copy-paste ready. | 2 / 3 |
Workflow Clarity | The workflow is exceptionally well-sequenced with a clear 4-step mandatory cycle (task-executor → escalation check → quality-fixer → commit), explicit validation checkpoints at each step, feedback loops for error recovery (stub_detected → return to step 2, needs_revision → return with fixes), and clear escalation paths for blocked states. The decision tables for task generation and implementation readiness are well-structured. | 3 / 3 |
Progressive Disclosure | The skill references several external skills and subagents (subagents-orchestration-guide, task-decomposer, quality-fixer, code-verifier, security-reviewer, document-reviewer) but no bundle files are provided. The content itself is organized into clear sections with good headers, but the monolithic nature of the file—with repeated Consumed Task Set computation rules and lengthy pre-execution checks—suggests some content could be split into referenced files. | 2 / 3 |
Total | 9 / 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 |
|---|---|---|
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 10 / 11 Passed | |
- Repository
- shinpr/claude-code-workflows
- Commit
adf2e4d
- 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.