help · yonatangross — Claude Code skill: what it does, how to install
This skill scans the installed OrchestKit skills and sorts them into categories. It helps you find the workflow or skill that matches a task.
MIT 166 installs ★ 305 updated 2026-11-07
npx skills add yonatangross/orchestkit --skill help -a claude-codeWhat help does
help builds a live directory of the OrchestKit skills available in the current installation. It reads each matched file’s frontmatter, then collects its name, description, version, complexity, argument hint, and tags. The results are grouped into areas such as build, git, plan, quality, memory, config, explore, design, and ops.
With no argument, the skill asks you to choose a broad task group through AskUserQuestion. Passing build, git, or all selects a listing mode directly. Discovery uses Grep, Read, and Glob; the installed plugin is checked through ${CLAUDE_PLUGIN_ROOT}/skills, with src/skills used as the repository fallback. Claude Code 2.1.277+ is required.
Each entry includes its command, version, complexity, description, and an argument example. The all mode prints every user-invocable skill and derives the final count from the files it found.
What it helps with
- find a skill by task category
- list the OrchestKit commands available in the installation
- filter skills for
buildorgitwork - show every user-invocable skill
- compare a task with a skill description and example
Example prompts for the agent
- Run
/ork:helpand show the available skill categories. - Run
/ork:help buildand list skills for implementation work. - Run
/ork:help alland print every user-invocable skill.
How to install help
1Skills CLI (Vercel)
npx skills add yonatangross/orchestkit --skill help -a claude-code2Manually: put the files in place
~/.claude/skills/help/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/help/
1Skills CLI (Vercel)
npx skills add yonatangross/orchestkit --skill help -a codex2Manually: put the files in place
.agents/skills/help/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add yonatangross/orchestkit --skill help -a cursor2Manually: put the files in place
.cursor/skills/help/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install yonatangross/orchestkit src/skills/help2Manually: put the files in place
.github/skills/help/SKILL.md1Skills CLI (Vercel)
npx skills add yonatangross/orchestkit --skill help -a gemini-cli2Platform CLI
gemini skills install https://github.com/yonatangross/orchestkit --path src/skills/helpManually: put the files in place
.devin/skills/help/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add yonatangross/orchestkit --skill help -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about help
How do I install help in Claude Code?
Run npx skills add yonatangross/orchestkit --skill help -a claude-code, then use ~/.claude/skills/help/SKILL.md.
What does the help skill do?
It discovers the installed OrchestKit skills at runtime, groups them by category, and displays their descriptions, versions, complexity, and invocation examples.
Which Claude Code versions support help?
The skill supports Claude Code 2.1.277+.
Source text
SKILL.md · MIT
---
name: help
license: MIT
compatibility: "Claude Code 2.1.277+."
description: "OrchestKit skill directory by category. Use when looking for the right skill or workflow."
argument-hint: "[category]"
context: inherit
user-invocable: true
allowed-tools: "AskUserQuestion Read Grep Glob"
effort: low
model: haiku
metadata:
category: document-asset-creation
version: "2.1.0"
author: "OrchestKit"
complexity: "low"
tags: "help, documentation, skills, discovery, meta"
triggers:
keywords: [help, "what skills", "skills can i", "available commands", "which skill", "show me skills", "ork commands", "what can", "whats ork", "skill for"]
examples:
- "what orchestkit skills can i use"
- "show me the available ork commands"
- "help me find the right skill for code review"
anti-triggers: [fix, implement, doctor, setup, explore, assess]
---
# OrchestKit Skill Directory
Dynamic skill discovery: enumerates the installed plugin at runtime so listings are never stale.
> **CC 2.1.121+ tip:** if you just want to find one skill quickly, the native `/skills` command now has type-to-filter — open it and start typing the skill name. Use `help` when you want categorized browsing or rationale for *why* a skill applies.
## Quick Start
```bash
help # Show all categories
help build # Show BUILD skills only
help git # Show GIT skills only
help all # List every user-invocable skill
```
## Argument Resolution
```python
CATEGORY = "$ARGUMENTS[0]" # Optional: build, git, plan, quality, memory, config, explore, design, ops, all
# If provided, skip AskUserQuestion and show that category directly.
# $ARGUMENTS is the full string (CC 2.1.59 indexed access)
```
---
## STEP 0: Dynamic Skill Discovery
**ALWAYS run this first** to get accurate, up-to-date skill data:
```python
# ${CLAUDE_PLUGIN_ROOT} is set by the plugin runtime and points at the INSTALLED
# plugin. That is the normal case: a marketplace user has no src/ directory.
SKILLS_ROOT = "${CLAUDE_PLUGIN_ROOT}/skills"
matches = Grep(pattern="user-invocable:\\s*true", path=SKILLS_ROOT, output_mode="files_with_matches")
# Dogfooding fallback: inside the OrchestKit repo itself the skills live in the
# source tree. Retry there if the env var was unset or the probe found nothing.
if not matches:
SKILLS_ROOT = "src/skills"
matches = Grep(pattern="user-invocable:\\s*true", path=SKILLS_ROOT, output_mode="files_with_matches")
```
If BOTH probes return zero files, say so plainly ("could not locate the OrchestKit
skills directory, checked `${CLAUDE_PLUGIN_ROOT}/skills` and `src/skills`") and stop.
Do NOT substitute a remembered list of skill names. Any such list is stale by
construction and a confident wrong answer is worse than no answer.
For each matched file, read the frontmatter to get name, description, version,
complexity, `argument-hint` and tags:
```python
Read(file_path=f"{SKILLS_ROOT}/{skill_dir}/SKILL.md", limit=25)
```
Every number rendered later is derived from this scan, never typed as a literal:
```python
TOTAL = len(matches) # user-invocable skill count
```
Build the skill list dynamically. **Never hardcode counts or skill names.**
---
## STEP 1: Category Selection
If CATEGORY argument provided, skip to STEP 2 with that category.
Otherwise, present categories interactively:
```python
AskUserQuestion(
questions=[{
"question": "What type of task are you working on?",
"header": "Category",
# 4-option cap (CC schema): every category from STEP 2 is grouped into one of
# 3 buckets + "Show all". STEP 2 renders the constituent categories for the
# picked bucket. Descriptions name intents, never skills, so they cannot drift.
"options": [
{"label": "Build & ship", "description": "Writing code, tests, git and PRs, UI and design work"},
{"label": "Plan & assess", "description": "Requirements, planning, quality assessment, review"},
{"label": "Explore & operate", "description": "Codebase exploration, memory, setup, diagnostics, CI"},
{"label": "Show all", "description": "List every user-invocable skill"}
],
"multiSelect": false
}]
)
```
---
## STEP 2: Render Category
For the selected category, render the skill table from the data gathered in STEP 0.
### Category Definitions
Categories are defined by **tag predicates**, never by a list of skill names, so a
newly shipped skill lands in the right bucket without editing this file. Match each
discovered skill's frontmatter `tags` against the sets below (case-insensitive, one
hit is enough):
| Category | CLI arg | Matches any of these tags |
|----------|---------|---------------------------|
| BUILD | `build` | implementation, feature, testing, coverage, test-generation, verification, e2e |
| GIT | `git` | git, github, commit, pull-request, pr, issue, bug-fix |
| PLAN | `plan` | planning, ideation, prd, requirements, visualization |
| QUALITY | `quality` | quality, assessment, evaluation, code-review, validation, grading |
| MEMORY | `memory` | memory, decisions, graph-memory, consolidation |
| CONFIG | `config` | setup, configuration, onboarding, diagnostics, health-check, dev-loop |
| EXPLORE | `explore` | exploration, codebase, code-search, architecture, discovery |
| DESIGN | `design` | design, design-context, design-tokens, design-to-code, frontend, ui, components, stylecards |
| OPS | `ops` | ci, automation, telemetry, observability, release, migration |
| OTHER | (none) | anything the rows above did not match |
A skill matching two categories is listed under both. That is expected, not a bug.
OTHER is what makes the totals reconcile: every skill found in STEP 0 must appear
somewhere in a full listing, so a skill nobody has categorized yet still shows up.
The STEP 1 picker only offers **3 buckets** (the AskUserQuestion schema caps a
question at 4 options). Each bucket renders the union of its categories:
| Picker bucket | Renders categories |
|---------------|--------------------|
| Build & ship | BUILD + GIT + DESIGN |
| Plan & assess | PLAN + QUALITY |
| Explore & operate | MEMORY + CONFIG + EXPLORE + OPS + OTHER |
For each skill in the category, render:
```
/ork:{name} v{version} {complexity}
{description}
Example: /ork:{name} {argument-hint example}
```
### "Show all" — Full Listing
If user picks "Show all", render ALL user-invocable skills grouped by category from
STEP 0 data, then close with the derived total: `{TOTAL} user-invocable skills`.
Print `TOTAL` from `len(matches)`. Never type a number you did not just count.
---
## CC Built-in Commands (2.1.72+)
Not OrchestKit skills — these are Claude Code built-ins:
| Command | Description | Since |
|---------|-------------|-------|
| `/code-review` | Review changed code for correctness bugs at a chosen effort level; `--comment` posts inline PR comments (renamed from `/simplify` in 2.1.146; the old cleanup-and-fix behavior was removed) | CC 2.1.146 |
| `/help` | Claude Code built-in help | CC 2.1.0+ |
| `/config` | Claude Code configuration | CC 2.1.0+ |
| `/clear` | Clear conversation (preserves background agents) | CC 2.1.72 |
| `/fast` | Toggle fast mode (same model, faster output) | CC 2.1.59+ |
| `/loop` | Recurring interval (e.g. `/loop 5m /foo`) | CC 2.1.71 |
| `/plan` | Enter plan mode (CC 2.1.222 removed the ultraplan feature, so `/ultraplan` and the "Refine with Ultraplan" hand-off no longer exist) | CC 2.1.72 |
| `/team-onboarding` | Generate teammate ramp-up guide | CC 2.1.101 |
| `/recap` | Session context restoration after idle | CC 2.1.108 |
| `/undo` | Alias for `/rewind` | CC 2.1.108 |
| `/skills` | Native skill picker — type to filter | CC 2.1.121 |
| `/effort` | Reasoning effort — sets `$CLAUDE_EFFORT` env for skills | CC 2.1.72 (env var since 2.1.120) |
| `/cd` | Move session to a new working directory without breaking the prompt cache | CC 2.1.169 |
| `--safe-mode` | Start with ALL customizations disabled (troubleshooting bisect) | CC 2.1.169 (flag/env, not a slash command) |
## Performance Tips
- **`ENABLE_PROMPT_CACHING_1H=1`** — Extends prompt cache TTL from 5 min to 1 hour. Set this for long sessions with multi-phase skills (brainstorm, implement, cover). Available on API key, Bedrock, Vertex, Foundry.
- **`/recap`** — Restores session context after stepping away. Complements OrchestKit's checkpoint-resume for chain recovery.
## Keyboard Shortcuts
| Shortcut | Action |
|----------|--------|
| `Ctrl+F` | Find in session output |
| `Esc` | Cancel / dismiss |
| `Shift+Enter` | Newline in input |
| `Ctrl+C` | Cancel operation |
---
## Pro Tip
You don't need to memorize skills. Just describe your task naturally:
```
"I need to implement user login" → implement
"Show me the payment architecture" → explore
"Review PR 123" → review-pr
"Is this code good?" → assess
"Plan out the billing redesign" → visualize-plan
```
## Related Skills
- `/help` — Claude Code built-in help
- `doctor` — OrchestKit health check
- `setup` — Full onboarding wizard
© OrchestKit · MIT · yonatangross/orchestkit
Similar skills
All skills for Agent setup: memory, context, orchestration →help by other authors comes first.