repo-skill-authoring
Create and update repo-local skills that follow the MetaMask Mobile standard. Use for new repo-local skills, skill entrypoints, or harness shims in this repository.
73
66%
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 ./.agents/skills/repo-skill-authoring/SKILL.mdQuality
Discovery
75%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 is well-structured with a clear 'what' and 'when' clause, and targets a very specific niche (MetaMask Mobile repo-local skill creation). Its main weakness is that the trigger terms are quite technical and project-specific, and the concrete actions could be more detailed (e.g., what does creating a skill entrypoint or harness shim involve?).
Suggestions
Add more natural trigger terms a user might say, such as 'add a new skill', 'scaffold skill', 'create skill file', or 'write a .skill.md'.
List more specific concrete actions beyond 'create and update', e.g., 'generate YAML frontmatter, write skill markdown files, create test harness shims, define entrypoint configurations'.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Names the domain (repo-local skills for MetaMask Mobile) and some actions (create, update), but doesn't elaborate on what creating/updating entails or list comprehensive concrete actions beyond mentioning 'skill entrypoints' and 'harness shims'. | 2 / 3 |
Completeness | Clearly answers both 'what' (create and update repo-local skills following MetaMask Mobile standard) and 'when' ('Use for new repo-local skills, skill entrypoints, or harness shims in this repository'), with explicit trigger guidance. | 3 / 3 |
Trigger Term Quality | Includes some relevant terms like 'repo-local skills', 'skill entrypoints', 'harness shims', and 'MetaMask Mobile', but these are fairly technical/project-specific. Missing more natural user phrases like 'add a new skill', 'write a skill file', or 'scaffold a skill'. | 2 / 3 |
Distinctiveness Conflict Risk | Highly specific niche — repo-local skills for MetaMask Mobile with mentions of skill entrypoints and harness shims. Very unlikely to conflict with other skills due to the narrow, project-specific scope. | 3 / 3 |
Total | 10 / 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 competent meta-skill for repo-local skill authoring that provides good structural guidance and useful minimal examples. Its main weaknesses are moderate verbosity (some concepts are restated across sections) and the Agent Execution Standard steps being more descriptive than prescriptive—they tell you what to think about rather than giving exact commands or validation procedures. The progressive disclosure and organization are strong.
Suggestions
Make the 'Sanity check the result' step more concrete—provide a specific command or checklist (e.g., 'grep for the canonical path in all shim files' or a validation script) rather than just 'Confirm the skill and any shims all point to the same workflow'.
Demonstrate the 'Required agent response sections' (Implementation Checklist, Files To Add Or Modify, Validation, Assumptions) with a filled-in example so Claude knows exactly what format to produce.
Consolidate the 'When To Use This vs Native Skill Creator' and 'Harness Entrypoints' sections into the Purpose section to reduce redundancy and tighten the overall token budget.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is reasonably structured but includes some redundant sections (e.g., 'When To Use This vs Native Skill Creator' and 'Harness Entrypoints' could be tighter). Some guidance like 'Keep the repo skill as the canonical artifact' is repeated in different forms across sections. The 'do not create a skill when' list is useful but the overall content could be trimmed by ~20%. | 2 / 3 |
Actionability | The skill provides concrete directory structures and minimal examples (SKILL.md template, Claude shim, openai.yaml), which are helpful. However, the 'Agent Execution Standard' steps are somewhat abstract ('Define the user/problem shape', 'Split the content deliberately') rather than providing executable commands or copy-paste-ready templates with filled-in values. The required response sections are listed but not demonstrated. | 2 / 3 |
Workflow Clarity | The 5-step Agent Execution Standard provides a clear sequence, and step 5 ('Sanity check the result') serves as a validation checkpoint. However, the validation step is vague—it says 'Confirm the skill and any shims all point to the same workflow' without specifying how to verify this (e.g., a script, a grep command, or a checklist). For a skill that involves creating files across multiple directories, more explicit validation would be expected. | 2 / 3 |
Progressive Disclosure | The skill is well-organized with clear sections (Purpose, When To Use, Harness Entrypoints, Repo Skill Shape, Naming Rules, Agent Execution Standard, Minimal Examples). It appropriately keeps the overview in the SKILL.md and mentions scripts/ and references/ as optional directories for detailed content. The structure is flat and easy to navigate. | 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
- MetaMask/metamask-mobile
- Commit
bee9b14
- 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.