feat(axi): add AXI skill plugin to marketplace#228
Conversation
Add the AXI (Agent eXperience Interface) skill plugin, which provides ergonomic standards and conventions for building agent-facing CLI tools. Sourced from github.com/kunchenguid/axi via skills.sh (bunx skills add). - .claude-plugin/marketplace.json: added axi entry (source of truth) - .agents/plugins/marketplace.json: regenerated Codex marketplace with axi - plugins/axi/: new plugin directory with skill files and skills-lock.json - plugins/axi/.claude-plugin/plugin.json: Claude Code manifest - plugins/axi/.codex-plugin/plugin.json: Codex manifest (multi-format) - plugins/axi/plugin.json: Antigravity manifest (multi-format) - release-please-config.json: added plugins/axi package with 3 manifests - .release-please-manifest.json: added plugins/axi version 1.0.0 - README.md: added AXI entry under Built-in Plugins
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
|
Caution Review failedThe pull request is closed. ℹ️ Recent review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (10)
📝 WalkthroughWalkthroughAdds a new Changesaxi Plugin Addition
Estimated code review effort🎯 2 (Simple) | ⏱️ ~10 minutes Possibly related PRs
Poem
✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
🔍 Tessl Skill Review
|
| Dimension | Score | Detail |
|---|---|---|
| conciseness | ██░ 2/3 | The skill is reasonably well-written but contains some verbosity — explanatory sentences like 'The most expensive token cost is often not a longer response — it's a follow-up call' and rationale paragraphs before each rule add tokens without teaching Claude anything new. Several sections could be tightened by removing motivational framing and keeping only the rules and examples. |
| actionability | ███ 3/3 | The skill provides concrete, copy-paste-ready output examples for nearly every principle (TOON format, truncation patterns, error structures, empty states, session hooks, contextual disclosure). The examples are specific enough that Claude can directly apply them when building CLI tools. Integration instructions for Claude Code, Codex, and OpenCode include specific file paths and configuration details. |
| workflow clarity | ██░ 2/3 | The 10 numbered sections provide a clear organizational structure, but there's no explicit sequencing for building an AXI-compliant CLI — the sections read more as independent principles than a workflow. For a skill about building tools, there are no validation checkpoints (e.g., 'verify your output is valid TOON', 'test that empty states are definitive'). The session integration section (7) describes a multi-step process but lacks explicit verification steps. |
| progressive disclosure | ██░ 2/3 | The skill references the external TOON specification appropriately, but the body itself is quite long (~300+ lines) with all content inline. Sections like the detailed session integration rules for each app (Claude Code, Codex, OpenCode) and the installable skill subsection could be split into separate reference files. No bundle files are provided to offload this content. |
Overall: This is a well-structured, highly actionable skill with excellent concrete examples for each principle. Its main weaknesses are verbosity (motivational/rationale sentences before each rule that Claude doesn't need) and the monolithic structure — at this length, some sections should be split into referenced files. The lack of explicit build/validation workflow steps is a minor gap given the skill's nature as a design standard rather than a step-by-step procedure.
Suggestions:
- Remove rationale/motivation sentences before each section (e.g., 'The most expensive token cost is often not a longer response...') — Claude doesn't need persuading, just the rules and examples.
- Split the lengthy session integration section (7) into a separate INTEGRATIONS.md file, keeping only a summary and link in the main skill.
- Add a brief validation checklist at the end (e.g., 'Verify: no-args shows content not help, empty states are explicit, errors go to stdout, no interactive prompts') to provide a concrete review workflow.
To improve your score, point your agent at the Tessl optimization guide. Need help? Jump on our Discord.
Feedback
Report issues with this review at tesslio/skill-review, or send private feedback from your terminal with tessl feedback.
Up to standards ✅🟢 Issues
|
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request introduces the Agent eXperience Interface (AXI) plugin, adding its configuration files, release configurations, README documentation, and the AXI skill specification. The reviewer provided feedback on .claude-plugin/marketplace.json pointing out that the newly added axi plugin entry lacks category, keywords, and tags fields, which are standard for other marketplace entries, and suggested adding them to maintain consistency.
| { | ||
| "name": "axi", | ||
| "description": "Agent eXperience Interface (AXI) — ergonomic standards for building CLI tools that agents use via shell execution. Use when building, modifying, or reviewing any agent-facing CLI.", | ||
| "source": "./plugins/axi" | ||
| } |
There was a problem hiding this comment.
[MEDIUM] AXI 플러그인 메타데이터 누락
Problem: .claude-plugin/marketplace.json 파일의 axi 플러그인 항목에 category, keywords, tags 필드가 누락되어 다른 플러그인 항목들과의 일관성이 결여되어 있습니다.
Rationale: 마켓플레이스 메타데이터의 일관성 및 유지보수성(Organization Style Guide Key Principles - Maintainability). 해당 필드가 누락되면 마켓플레이스 내에서 카테고리 분류나 검색 기능이 정상적으로 작동하지 않을 수 있습니다.
Suggestion: 다음과 같이 category, keywords, tags 필드를 추가해 주세요.
{
"name": "axi",
"description": "Agent eXperience Interface (AXI) — ergonomic standards for building CLI tools that agents use via shell execution. Use when building, modifying, or reviewing any agent-facing CLI.",
"category": "productivity",
"keywords": ["axi", "cli", "agent-experience", "standards"],
"tags": ["skills", "standards"],
"source": "./plugins/axi"
}References
- Organization Style Guide Key Principles - Maintainability: Ensure future engineers can read and modify it easily, and maintain consistency across the codebase. (link)
|
1 participant
Summary
Add the AXI (Agent eXperience Interface) skill plugin to the claude-code-plugins marketplace.
AXI provides ergonomic standards and conventions for building agent-facing CLI tools — covering output formatting, exit codes, error messaging, and other interface contracts that make CLI tools pleasant to use from agents.
How it was generated
Changes
.claude-plugin/marketplace.json: added axi entry (source of truth).agents/plugins/marketplace.json: regenerated Codex marketplace with axiplugins/axi/: new plugin directory with SKILL.md and skills-lock.jsonplugins/axi/.claude-plugin/plugin.json: Claude Code manifestplugins/axi/.codex-plugin/plugin.json: Codex manifest (auto-generated)plugins/axi/plugin.json: Antigravity manifest (auto-generated)release-please-config.json: added plugins/axi package with 3 version-bearing manifests.release-please-manifest.json: added plugins/axi at version 1.0.0README.md: added AXI entry under Built-in PluginsTest Plan
/plugin install axi@pleaseaiSummary by cubic
Adds the
axiskill plugin to the marketplace, providing ergonomic standards for agent‑facing CLI tools. Includes Claude Code, Codex, and Antigravity manifests withrelease-pleaseset to v1.0.0 forplugins/axi.plugins/axi/with skill content and manifests (.claude-plugin/plugin.json,.codex-plugin/plugin.json,plugin.json).axiin.claude-plugin/marketplace.jsonand.agents/plugins/marketplace.json; documented install inREADME.md.release-please-config.jsonand.release-please-manifest.jsonto manageplugins/axireleases at1.0.0.Written for commit af82724. Summary will update on new commits.
Summary by CodeRabbit
Release Notes
New Features
Documentation