New: 98,000+ skills indexed from GitHub repos with 100+ stars · every skill safety-checkedBrowse top skillsTop skills
Esc to closeBrowse all skills
Submit a skillBrowse skills
Submit a skillBrowse skills

Official agent skill

Review Huabu Agent

by microsoft in microsoft/Huabu

Review the Huabu operate agent's three steering artifacts at their fixed repo locations — system prompt, tool descriptions, and skills.

OfficialMITAuto-check passedAI & LLM Engineering

Install Review Huabu Agent

skills CLI
$ npx skills add microsoft/Huabu --skill review-huabu-agent -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install microsoft/Huabu review-huabu-agent --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/microsoft/Huabu.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/review-huabu-agent .claude/skills/review-huabu-agent && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
review-huabu-agent
GitHub stars
157
Token cost
~1.2k tokens
SKILL.md length
387 words
Files
1
Skills in repo
10
Repo updated
First seen
Licence
MIT

At a glance

Review the Huabu operate agent's three steering artifacts at their fixed repo locations — system prompt, tool descriptions, and skills.

  • Works in 4 steps: Read the general rubric first (link… → Default to a combined review of all… → Score with the shared foundation + the… → …
  • Polishing the operate (or ask/intent/sketch) agent prompt
  • SKILL.md covers Artifact locations, How to review and Huabu cross-artifact checks
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Review Huabu Agent is an agent skill from microsoft/Huabu, published by the product's own GitHub organization. Review the Huabu operate agent's three steering artifacts at their fixed repo locations — system prompt, tool descriptions, and skills. USE WHEN reviewing, auditing, or polishing the operate (or ask/intent/sketch) agent prompt, the agent tool definitions, or the canvas/memory skills; before shipping changes to AGENT.md, tool definitions.ts, or a skill. Thin Huabu-specific wrapper: it only locates the artifacts; all scoring rules come from review-agent-primitives.

Its SKILL.md is about 1.2k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in AI & LLM Engineering, covering Structured output and tool calling and Prompt engineering. The repository describes itself as: Huabu, where you and your agents think together. The licence is MIT.

When your agent uses it

  • Polishing the operate (or ask/intent/sketch) agent prompt
  • The agent tool definitions
  • The canvas/memory skills
  • Before shipping changes to AGENT.md

Example prompts

  • “/review-huabu-agent”

Workflow steps

4 steps, taken from the first numbered list in SKILL.md.

  1. Read the general rubric first (link above), then read the artifact(s) being reviewed at the paths above.
  2. Default to a combined review of all three so cross-artifact and placement issues surface; do a single-artifact pass only if the user…
  3. Score with the shared foundation + the matching reference rubric; run Placement mode whenever content looks misplaced.
  4. Report using review-agent-primitives' output format (verdict → findings table → prioritized fixes → optional rewrite).

What it can do on your machine

Read from SKILL.md and the folder at commit ef48105. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Review Huabu Agent loads about 1.2k tokens when it runs. Until then it costs about 122 tokens; SKILL.md has 387 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~122
When it runs · the whole SKILL.md, loaded when a task matches
~1.2k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from microsoft/Huabu at commit ef48105, republished under its MIT licence (© microsoft). 387 words, ~1,171 tokens.

Download SKILL.mdSave it as .claude/skills/review-huabu-agent/SKILL.md (or your agent's skills folder).
name
review-huabu-agent
description
Review the Huabu operate agent's three steering artifacts at their fixed repo locations — system prompt, tool descriptions, and skills. USE WHEN reviewing, auditing, or polishing the operate (or ask/intent/sketch) agent prompt, the agent tool definitions, or the canvas/memory skills; before shipping changes to AGENT.md, tool definitions.ts, or a skill. Thin Huabu-specific wrapper: it only locates the artifacts; all scoring rules come from review-agent-primitives.

Review Huabu Agent

Huabu-specific entry point for reviewing this repo's agent primitives. It adds only the artifact locations. Every rubric, the scoring model, the output format, and the placement decision table live in the general skill — do not duplicate them here.

Rules & rubric: review-agent-primitives plus its references: tool-description · agent-prompt · skill-quality.

Artifact locations

ArtifactPathRubric to apply
System promptapps/server/src/prompt/agents/operate/AGENT.md (same structure for ask/, intent/, sketch/)agent-prompt
Tool descriptionsapps/server/src/modules/agent/tools/definitions.ts plus the canonical canvas command/query Zod schemas in packages/shared/src/types/api/space-operations.ts — their description strings are load-bearing text the model reads at call time, so review them together, not separatelytool-description
Skillsapps/server/src/prompt/skills/ — canvas/ (its SKILL.md and every file under canvas/references/), memory/, sketch-gestures/, create-skill/, update-skill/skill-quality

How to review

  1. Read the general rubric first (link above), then read the artifact(s) being reviewed at the paths above.
  2. Default to a combined review of all three so cross-artifact and placement issues surface; do a single-artifact pass only if the user scopes it that way.
  3. Score with the shared foundation + the matching reference rubric; run Placement mode whenever content looks misplaced.
  4. Report using review-agent-primitives' output format (verdict → findings table → prioritized fixes → optional rewrite).
Show full SKILL.md (196 more words)Show less

Huabu cross-artifact checks

When reviewing more than one artifact together, pay special attention to:

  • Tool descriptions in definitions.ts must not contradict the tool guidance in AGENT.md.
  • Multi-step canvas procedures belong in prompt/skills/, not restated inside AGENT.md.
  • Tools referenced by AGENT.md must exist and be scoped in definitions.ts (no dangling references).
  • Skill name/description under prompt/skills/ must not collide, and each must carry clear "use when / don't use when" triggers.
  • Same-fact fan-out (lockstep-sync smell). One rule about the canvas often lives in many same-type files at once — a definitions.ts tool desc and its schemas/*.ts field description, or a SKILL.md and its own references/*.md, or two sibling skills (canvas/ vs sketch-gestures/). The cross-artifact-type checks above will not catch this because both copies are the same type. When you find one, trace every other file that states the same fact (grep the distinctive phrase), verify they don't drift or contradict after a change, and flag it as a Warn: converge to one canonical owner with a full statement + pointers, unless the copies genuinely diverge by audience/path (e.g. the chat "omit id" vs sketch "set explicit id" split) — in which case keep them but note that they must change in lockstep.

© microsoft, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .agents/skills/review-huabu-agent of microsoft/Huabu.

Open the folder on GitHubat commit ef48105

Compare with similar skills

Review Huabu Agent next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Review Huabu Agent compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Review Huabu Agent this skillmicrosoft/Huabu157—~1.2kAutomated safety check: PassMIT
Prompt Engineering Patternswshobson/agents40k—~1.3kAutomated safety check: PassMIT
Agent Prompt Quality Barmastra-ai/mastra29k—~2kAutomated safety check: PassCustom licence
Kayba Stage 2 Domain Contextkayba-ai/agentic-context-engine2.6k—~1.9kAutomated safety check: PassApache-2.0
Lintlanghermes-labs-ai/lintlang140—~719Automated safety check: PassApache-2.0
Lintlang Audithermes-labs-ai/lintlang140—~1.9kAutomated safety check: PassApache-2.0

Similar skills

  • Prompt Engineering Patterns

    wshobson/agents

    Reference for designing and tuning production LLM prompts: few-shot examples, chain-of-thought, structured outputs, templates and system prompts.

    40k GitHub stars~1.3k tokensUpdated 5 days ago
    AI & LLM EngineeringAuto-check passed

  • Agent Prompt Quality Bar

    mastra-ai/mastra

    Universal quality bar and final audit rubric for any agent system prompt.

    29k GitHub stars~2k tokensUpdated today
    AI & LLM EngineeringAuto-check passed

  • Kayba Stage 2 Domain Context

    kayba-ai/agentic-context-engine

    Gather domain context about the repository and agent — system prompt, tool definitions, domain docs, and behavior patterns from traces.

    2.6k GitHub stars~1.9k tokensUpdated 16 days ago
    AI & LLM EngineeringAuto-check passed

  • Lintlang

    hermes-labs-ai/lintlang

    A skill your agent uses when writing or reviewing AI agent configs, system prompts, or tool definitions (JSON/YAML/Python) and you need to catch ambiguous tool descriptions, missing stop conditions…

    140 GitHub stars~719 tokensUpdated today
    AI & LLM EngineeringAuto-check passed

  • Lintlang Audit

    hermes-labs-ai/lintlang

    Audit a named AI agent config, system prompt, tool definition, or instruction file (YAML, JSON, Markdown, text, or Python) with the released LintLang CLI in GitHub Copilot CLI.

    140 GitHub stars~1.9k tokensUpdated today
    AI & LLM EngineeringAuto-check passed

  • Agents Workflows

    VectorSpaceLab/AREX-Skill

    A skill your agent uses for giskard.agents async chat workflows, tools, prompt templates, structured outputs, retries, rate limiting, embeddings, and optional LiteLLM backend.

    331 GitHub stars~500 tokensUpdated 1 mo ago
    AI & LLM EngineeringAuto-check passed

More from microsoft/Huabu

All 10 skills in this repo

  • Code Review Expert

    microsoft/Huabu

    Official

    Expert code review of current git changes with a senior engineer lens.

    157 GitHub starsUsed in 2 repos~1.5k tokens
    Auto-check passed

  • Release

    microsoft/Huabu

    Official

    Use ONLY when the user explicitly asks to cut/publish a Huabu desktop release or tag a version.

    157 GitHub stars~1.1k tokensUpdated today
    Auto-check passed

  • Review Agent Primitives

    microsoft/Huabu

    Official

    Review the quality of an agent's tool descriptions, system/agent prompts, or SKILL.md files against current agent-engineering best practices.

    157 GitHub stars~2.3k tokensUpdated today
    Auto-check passed

  • Space

    microsoft/Huabu

    Official

    Space mental model (the infinite work surface), tool boundaries, and command reference.

    157 GitHub stars~2.4k tokensUpdated today
    Auto-check passed

  • Issue Tracker

    microsoft/Huabu

    Official

    Coordinate one or more GitHub issues through isolated Git worktrees, durable Huabu Tasks, and dedicated Fixing Agent Threads.

    157 GitHub stars~279 tokensUpdated today
    Auto-check passed

  • Deepv Slides Maker

    microsoft/Huabu

    Official

    Create and revise editable slide decks, PowerPoint files, and slide images with DeepV.

    157 GitHub stars~276 tokensUpdated today
    Auto-check: notes

Categories

Questions about Review Huabu Agent

What does Review Huabu Agent do?

Review the Huabu operate agent's three steering artifacts at their fixed repo locations — system prompt, tool descriptions, and skills. Review Huabu Agent is an agent skill from microsoft/Huabu, published by the product's own GitHub organization. Review the Huabu operate agent's three steering artifacts at their fixed repo locations — system prompt, tool descriptions, and skills.

When should I use Review Huabu Agent?

Review Huabu Agent fits situations like: polishing the operate (or ask/intent/sketch) agent prompt; the agent tool definitions; the canvas/memory skills; before shipping changes to AGENT.md.

How do I install Review Huabu Agent in Claude Code?

Run `npx skills add microsoft/Huabu --skill review-huabu-agent -a claude-code`. Or copy the skill folder (.agents/skills/review-huabu-agent in microsoft/Huabu) into .claude/skills/review-huabu-agent in your project. Claude Code loads it when a task matches its description.

How do I install Review Huabu Agent in Codex?

Run `npx skills add microsoft/Huabu --skill review-huabu-agent -a codex`. Or copy the skill folder (.agents/skills/review-huabu-agent in microsoft/Huabu) into .agents/skills/review-huabu-agent in your project. Codex loads it when a task matches its description.

Can I use Review Huabu Agent in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add microsoft/Huabu --skill review-huabu-agent -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/review-huabu-agent, .gemini/skills/review-huabu-agent, .github/skills/review-huabu-agent and .opencode/skills/review-huabu-agent in your project.

What does Review Huabu Agent need to run?

SKILL.md names no scripts, command-line tools or credentials: Review Huabu Agent is instructions for the agent only.

Does Review Huabu Agent access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Review Huabu Agent safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Review Huabu Agent use?

Review Huabu Agent is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Review Huabu Agent use?

About 1.2k tokens (SKILL.md is roughly 4.7k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Review Huabu Agent?

Skills that share tags, products or a category with Review Huabu Agent: Prompt Engineering Patterns (wshobson/agents, 40k stars), Agent Prompt Quality Bar (mastra-ai/mastra, 29k stars), Kayba Stage 2 Domain Context (kayba-ai/agentic-context-engine, 2.6k stars) and Lintlang (hermes-labs-ai/lintlang, 140 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Review Huabu Agent?

microsoft (a GitHub organization, an official publisher) maintains it in microsoft/Huabu, which has 157 GitHub stars. The repository holds 10 skills in this directory. The repository was last updated on October 10, 2026.

Source: microsoft/Huabu on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.