spec-loop-implementation-flow
Use this skill after implementation approval for an approved chat-only or task-file task. It governs implementation-time routing, canonical task updates, and the final review checkpoint.
40
37%
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/spec-loop-implementation-flow/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 relies heavily on internal process jargon without explaining concrete actions or using natural trigger terms. While it attempts to specify a 'when' condition, the 'what' remains abstract and unclear. It would be very difficult for Claude to reliably select this skill from a pool of available skills.
Suggestions
Replace abstract jargon ('implementation-time routing', 'canonical task updates', 'final review checkpoint') with concrete actions describing what the skill actually does (e.g., 'Executes approved tasks, updates task status files, and performs final code review').
Add natural trigger terms that a user or system would use, such as 'implement task', 'start approved work', 'execute task', 'run implementation'.
Clarify the 'what' by listing specific operations performed during implementation, such as writing code, updating status files, running tests, or submitting for review.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description uses vague, abstract language like 'implementation-time routing', 'canonical task updates', and 'final review checkpoint' without describing any concrete actions. No specific operations are listed. | 1 / 3 |
Completeness | There is a 'when' clause ('after implementation approval for an approved chat-only or task-file task'), but the 'what' is extremely vague — 'routing', 'task updates', and 'review checkpoint' don't clearly explain what the skill actually does. | 2 / 3 |
Trigger Term Quality | The terms used ('implementation-time routing', 'canonical task updates', 'final review checkpoint') are internal jargon that no user would naturally say. There are no natural keywords a user would use to trigger this skill. | 1 / 3 |
Distinctiveness Conflict Risk | The description is so vague and jargon-heavy that it could easily overlap with any workflow or task management skill. Terms like 'routing', 'task updates', and 'review checkpoint' are generic process concepts. | 1 / 3 |
Total | 5 / 12 Passed |
Implementation
57%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-organized routing/dispatcher skill that effectively delegates detailed guidance to companion files with clear navigation. Its main weaknesses are that the workflow sequence could be more explicitly structured (numbered steps with checkpoints), and the skill itself contains limited directly actionable content since nearly all concrete guidance lives in referenced files. The content has minor redundancy but generally respects token efficiency.
Suggestions
Consolidate the workflow into a clear numbered sequence (e.g., 1. Verify approved task exists → 2. Read parent skill → 3. Select path → 4. Read companions → 5. Implement → 6. Run Implementation notes check → 7. Present for review) to improve workflow clarity.
Remove the duplicate 'Keep using the approved active task as the source of truth' instruction that appears both in Core rules and in the Read and path selection section.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The content is mostly efficient but has some redundancy—the instruction to 'keep the approved active task as the source of truth' appears twice, and several bullet points could be tightened. Some phrasing is slightly verbose (e.g., 'materially change scope, constraints, design, or test specification') but overall respects Claude's intelligence. | 2 / 3 |
Actionability | The skill provides concrete routing decisions and references to companion files, but the actual implementation guidance is deferred entirely to those companion files. The skill itself is more of a dispatcher than an actionable guide—it tells Claude what to read and which path to choose, but doesn't contain executable code, commands, or concrete examples of what implementation steps look like. | 2 / 3 |
Workflow Clarity | There is a discernible sequence (check for approved task → read parent skill → select path → read companions → implement → perform Implementation notes check → present work), but the steps are scattered across prose and bullet points rather than presented as a clear numbered workflow. Validation is mentioned ('Implementation notes check') but the actual validation steps are deferred to companion files, and there's no explicit feedback loop for error recovery within this skill. | 2 / 3 |
Progressive Disclosure | The skill is well-structured as an overview/dispatcher that clearly signals one-level-deep references to companion files (implementation-flow-guidance.md, chat-only-path-guidance.md, task-file-path-guidance.md) with explicit links and clear descriptions of what each file covers. The content appropriately splits concerns between the overview and detailed companions. | 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
- dpolivaev/spec-loop
- Commit
90f996e
- 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.