plan-task
Refine, parallelize, and verify a draft task specification into a fully planned implementation-ready task
50
40%
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 ./plugins/sdd/skills/plan-task/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.
The description uses process-oriented jargon that would not naturally match user queries, lacks a 'Use when...' clause, and provides only a high-level summary of what the skill does without concrete details. It would be difficult for Claude to reliably select this skill from a large pool because the trigger terms are too abstract and the scope is unclear.
Suggestions
Add an explicit 'Use when...' clause describing the scenarios that should trigger this skill, e.g., 'Use when the user has a draft task or feature spec that needs to be broken into parallel subtasks and validated before implementation.'
Replace abstract terms like 'refine', 'parallelize', and 'verify' with concrete actions users would recognize, such as 'break a task into parallel subtasks', 'validate dependencies', 'produce an implementation plan'.
Include natural trigger keywords a user might say, such as 'task planning', 'break down task', 'implementation plan', 'subtask', 'spec review'.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Names some actions ('refine', 'parallelize', 'verify') and a domain ('task specification'), but these terms are somewhat abstract. It doesn't list concrete sub-actions or outputs, and 'fully planned implementation-ready task' is vague about what that entails. | 2 / 3 |
Completeness | The description addresses 'what' at a high level but completely lacks a 'Use when...' clause or any explicit trigger guidance. Per the rubric, a missing 'Use when...' clause caps completeness at 2, and the 'what' is also weak, so this scores a 1. | 1 / 3 |
Trigger Term Quality | The terms used ('refine', 'parallelize', 'draft task specification', 'implementation-ready') are internal process jargon rather than natural keywords a user would say. Users are unlikely to say 'parallelize a task specification' or 'verify a draft task specification'. | 1 / 3 |
Distinctiveness Conflict Risk | The description is somewhat specific to task planning/specification workflows, which narrows the domain, but terms like 'refine' and 'verify' are generic enough to overlap with code review, editing, or other planning skills. | 2 / 3 |
Total | 6 / 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 highly actionable and well-structured orchestration workflow with excellent workflow clarity, explicit validation gates, and concrete executable guidance at every step. However, it is severely over-verbose — the ~800+ line monolithic file contains extensive inline content (rubrics, prompts, configuration logic, todo structures) that should be split into referenced files, and includes redundant explanations and repeated patterns that inflate token cost significantly.
Suggestions
Extract the 7 judge rubrics into a separate referenced file (e.g., `rubrics.md`) to reduce the main skill from ~800 lines to ~400 lines while preserving all actionable content.
Consolidate the repeated phase launch pattern (Description, Prompt, Capture, CRITICAL note) into a single template with a table of per-phase variations, eliminating ~200 lines of structural repetition.
Move the configuration resolution pseudocode and argument definitions into a separate `CONFIG.md` reference file — Claude can follow argument parsing without 100+ lines of explicit pseudocode.
Remove redundant CRITICAL warnings that repeat the same instruction (e.g., 'use prompt exactly as is' appears 7 times, 'DO NOT OUTPUT' appears in every prompt) — state once at the top and reference.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | This skill is extremely verbose at ~800+ lines. It over-specifies every detail including full JSON todo structures, complete prompt templates repeated for each phase, extensive configuration resolution pseudocode, and multiple redundant tables. Much of this could be condensed significantly — Claude doesn't need the level of hand-holding provided for argument parsing, todo filtering logic, or the repeated 'CRITICAL: use prompt exactly as is' warnings. | 1 / 3 |
Actionability | The skill provides highly concrete, executable guidance throughout: specific bash commands, exact agent prompts to use, precise file paths, complete rubric definitions with weights, and clear decision logic (PASS/FAIL thresholds). Every phase has copy-paste ready prompts and explicit capture requirements. | 3 / 3 |
Workflow Clarity | The multi-step workflow is exceptionally well-sequenced with explicit dependency chains, synchronization points, validation judges after every phase, retry flows with clear MAX_ITERATIONS limits, and error recovery paths. The ASCII workflow diagram, retry flow diagrams, and decision logic at each gate provide unambiguous sequencing with feedback loops. | 3 / 3 |
Progressive Disclosure | The skill is monolithic — all content is inline in a single massive file. The rubric templates, agent prompts, configuration resolution logic, and completion templates could be split into referenced files. It does reference external files like `${CLAUDE_PLUGIN_ROOT}/prompts/judge.md` and agent skill files, but the SKILL.md itself contains far too much inline content that should be separated (e.g., rubrics, prompt templates, todo JSON structures). | 2 / 3 |
Total | 9 / 12 Passed |
Validation
81%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 9 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
skill_md_line_count | SKILL.md is long (1224 lines); consider splitting into references/ and linking | Warning |
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 9 / 11 Passed | |
- Repository
- NeoLabHQ/context-engineering-kit
- Commit
dedca19
- 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.