Skills

A collection of skills for coding agents, focused on IndyKite — graph-based identity, authorization, and AI-agent integration. Manage your IndyKite project in the Hub UI (eu.hub.indykite.com / us.hub.indykite.com); developer docs at developer.indykite.com.

A skill is a self-contained bundle of instructions — and optionally scripts, references, or assets — that an agent can load on demand to perform a specialized task.

Available skills

Skill	What it does
indykite-agent-gateway	Deploy and configure Indykite Agent Gateway (IAG) in front of A2A agents to enforce caller, workflow, and delegation-chain checks.
indykite-mcp-server	Call the IndyKite MCP server (initialize session, list tools, call AuthZEN and ContX IQ tools) and configure the MCP server endpoint for a project.
indykite-ciq-read	Author a read-only IndyKite ContX IQ (CIQ) policy and Knowledge Query, then run it via POST /contx-iq/v1/execute.
indykite-ciq-create-node	Author an IndyKite CIQ policy + Knowledge Query that creates a brand-new node in the IKG, then run it via POST /contx-iq/v1/execute.
indykite-ciq-create-relationship	Author an IndyKite CIQ policy + Knowledge Query that creates a brand-new relationship between two existing nodes in the IKG, then run it via POST /contx-iq/v1/execute.
indykite-ciq-add-property	Author an IndyKite CIQ policy + Knowledge Query that sets one or more properties on an existing node in the IKG, then run it via POST /contx-iq/v1/execute.
indykite-ciq-add-relationship-property	Author an IndyKite CIQ policy + Knowledge Query that sets one or more properties on an existing relationship in the IKG, then run it via POST /contx-iq/v1/execute.
indykite-ciq-delete	Author an IndyKite CIQ policy + Knowledge Query that deletes a node, a relationship, or one or more properties from the IKG, then run it via POST /contx-iq/v1/execute.
indykite-ciq-create-node-with-link	Author an IndyKite CIQ policy + Knowledge Query that creates a brand-new node AND links it to one or more existing nodes via new relationships in a single atomic execute.

Examples gallery

Sample prompts each skill is designed to handle. Derived from each skill's description and ## When to use — they describe the kinds of asks the skill targets. Activation isn't guaranteed: which skill actually fires depends on the agent's matching algorithm, the model, and which other skills are installed. Verify routing in your own setup before relying on it.

Skill	Example prompt
indykite-agent-gateway	"Deploy IAG in front of my three A2A agents and wire up the workflow in the IKG."
indykite-mcp-server	"How do I initialise an MCP session against eu.mcp.indykite.com and call authzen_evaluate?"
indykite-ciq-read	"Expose Person-OWNS-Car as a parameterised read query through ContX IQ."
indykite-ciq-create-node	"Create a new Track node in the IKG with title and loudness, via CIQ."
indykite-ciq-create-relationship	"Add a PLAYED_AT relationship between an existing Track and an existing Venue."
indykite-ciq-create-node-with-link	"Create a new Contract and atomically link it to an existing Vehicle and Person."
indykite-ciq-add-property	"Let a Person update their own music_mood property."
indykite-ciq-add-relationship-property	"Annotate an existing PLAYED_AT relationship with a verified flag and timestamp."
indykite-ciq-delete	"Clear the music_mood property from a Person — GDPR-style erase."

Supported agents

These skills are installed via the skills CLI (skills.sh / vercel-labs/skills) — a cross-agent installer. One command, many agents: the CLI knows the per-agent install location for every agent it supports, and you target one (or all of them) via the --agent flag.

# Install into every agent the CLI detects on your machine
npx skills add <repo>

# Install for a specific agent
npx skills add <repo> --agent claude-code

# See which agent names the CLI accepts in your installed version
# (a typo prints the full valid set, so you discover it on demand)
npx skills add <repo> --agent foobar

The CLI accepts many agent names — far more than this repo has been tested against. Treat the CLI's own --agent list as the source of truth; it stays current as new agents are added.

What we've verified end-to-end in this repo:

• Claude Code — npx skills add <repo> --agent claude-code installs each skill into ~/.claude/skills/<name>/ (global) or .claude/skills/<name>/ (project). Claude Code reads SKILL.md directly and activates skills automatically by matching the user's prompt against description. There is also a bundle-install path via the Claude Code plugin marketplace using .claude-plugin/ — see Bundle install.

For Gemini CLI, install via the gemini-extension.json at the repo root instead — Gemini's extension system has its own loader, independent of the skills CLI.

For every other agent the skills CLI lists, npx skills add <repo> --agent <name> should drop the files in the right place — but native SKILL.md triggering is a property of the agent, not the CLI. The CLI guarantees files arrive at the right path; whether the agent then activates a skill automatically (by matching the prompt against description), via a slash command, or by explicit selection depends on that agent. Check the agent's own documentation when automatic activation matters.

Quickstart

The fastest way from zero to a skill activating in your agent:

# 1. Install one skill from this repo (project-scoped, into the right per-agent location)
npx skills add indykite/skills --skill indykite-mcp-server --agent claude-code

# 2. Restart the agent so it picks up the new skill directory.

# 3. Verify it loaded
npx skills list

Then, in the agent, ask something the skill's description matches — for indykite-mcp-server, that is anything about initializing an MCP session against eu.mcp.indykite.com / us.mcp.indykite.com, calling authzen_evaluate / ciq_execute, or debugging a 401 from the MCP server. The agent should pick the skill up automatically; if it doesn't, see How skills activate below.

To install all the skills in this repo at once, drop the --skill and --agent flags:

npx skills add indykite/skills

Installing

The recommended path is the skills CLI, which installs the same directory into the right location for whichever agent(s) you use:

# All skills in this repo, into the project's local agent directories
npx skills add indykite/skills

# All skills, globally for the current user
npx skills add indykite/skills -g

# Just one skill, into one agent
npx skills add indykite/skills --skill indykite-mcp-server --agent claude-code

# Show what's in the repo without installing
npx skills add indykite/skills --list

Useful flags:

• -g, --global — install at user scope instead of project scope.
• -a, --agent <name> — limit to one or more agents (use * for all).
• -s, --skill <name> — limit to one or more skills.
• --copy — copy files instead of symlinking (the default is a symlink so updates propagate).
• -y, --yes — skip confirmation prompts (handy in CI).

Restart the agent (or reload its config) after installing so it picks up the new skill directory.

Manual installation

If you don't want the CLI, copy or symlink the skill directory into the agent's expected location.

For Claude Code, that location is ~/.claude/skills/<skill-name>/ (user scope) or .claude/skills/<skill-name>/ (project scope) — Claude Code reads SKILL.md directly from there.

For any other agent, consult that agent's own documentation for where it expects rules, instructions, or skills files. Most agents that don't read SKILL.md natively expect the body of the file (everything after the YAML frontmatter) pasted into their own rule format.

Bundle install (Claude Code plugin / Gemini extension)

For agents that support a single labelled plugin install, this repo ships per-agent manifest files. They register all 9 skills at once and prompt the user for credentials (API_URL, API_KEY, BEARER_TOKEN, SERVICE_ACCOUNT_TOKEN, MCP_URL, PROJECT_GID) at install time.

Claude Code

The repo ships .claude-plugin/plugin.json and .claude-plugin/marketplace.json. Install via Claude Code's plugin marketplace in two steps:

# 1. Register this repo as a marketplace
/plugin marketplace add indykite/skills

# 2. Install the plugin from it
/plugin install indykite-skills

Gemini CLI

The repo ships gemini-extension.json at the root. Install via Gemini CLI's extension command:

gemini extensions install https://github.com/indykite/skills

Gemini reads the extension manifest, uses contextFileName (the repo's README.md) as loaded context, and exposes each entry in settings as an environment variable. The six configuration values (API_URL, API_KEY, BEARER_TOKEN, SERVICE_ACCOUNT_TOKEN, MCP_URL, PROJECT_GID) are the same as the Claude Code userConfig schema, sharing one source of truth in manifest.yaml.

Per-skill install via the skills CLI

If you only want one or two of the skills, or your agent doesn't have a plugin marketplace, use npx skills add indykite/skills --agent <name> — see Supported agents above. The bundle install registers all 9 skills with credential prompts; the per-skill install lets you cherry-pick and leaves credentials to env vars.

How skills activate

A skill is passive until invoked. Most supported agents activate skills automatically by matching the user's prompt against each installed skill's description field — the one-line summary in SKILL.md's frontmatter. When the description fits the request, the agent loads the rest of the skill into context before answering. The ## When to use section in each SKILL.md is read by the agent (not just the human) and is what determines whether activation actually fires for a given prompt — that is why the existing skills here list both positive triggers and explicit anti-triggers.

If you expect a skill to activate but it doesn't, check three things:

1. The agent loaded it. Ask the agent which skills are available, or run npx skills list (or npx skills ls) to see what is installed for the current scope. A skill that isn't installed cannot activate.
2. The description fits the prompt. Open SKILL.md and re-read the description and ## When to use. Vague descriptions get vague triggering — sharpen the wording, reinstall, and try again. If two skills could plausibly match the same prompt, the agent will pick one and you may need to disambiguate by tightening one of the descriptions.
3. Manual fallback. Most agents let you invoke a skill by name (e.g. /<skill-name> in Claude Code, or selecting it explicitly in Cursor / Copilot). Use that when automatic routing is uncertain — it also tells you whether the skill itself is loaded and working independently of the description match.

To disable a misbehaving skill: remove the directory (npx skills remove <name>, or delete it from the agent's skills folder), or set metadata.internal: true in its frontmatter so it stays installed but is hidden from automatic discovery (it will only surface if INSTALL_INTERNAL_SKILLS=1 is set).

Authoring a skill

Reference for anyone writing or forking a skill in this repo. The full submission flow lives in CONTRIBUTING.md — this section is the at-a-glance summary.

Structure

Each skill lives in its own directory with a SKILL.md file:

/
├── skill-name/
│   ├── SKILL.md
│   ├── scripts/      (optional)
│   ├── references/   (optional)
│   └── assets/       (optional)
└── another-skill/
    └── SKILL.md

SKILL.md format

---
name: your-skill-name
description: What it does. Use when [trigger conditions].
---

# Your Skill Name

## Instructions

### Step 1: ...
### Step 2: ...

Conventions

• Folder names are kebab-case (e.g. my-cool-skill).
• The file must be named exactly SKILL.md (case-sensitive).
• The description should state both what the skill does and when to invoke it — agents use it to decide whether the skill is relevant.
• Keep SKILL.md focused on core instructions; put longer docs in references/ and helper code in scripts/.

Contributing & security

• Want to add a new skill, fix one, or change the conventions? See CONTRIBUTING.md — it covers the quality bar, style, testing checklist, Agent Skills specification compliance, the submission process, and the code of conduct.
• Want to exercise the skills (structural validation, dry-run smoke tests, or live API roundtrips)? See testing/README.md. The runnable harness is ./testing/e2e-ciq.sh.
• Found a security issue (a skill that produces unsafe instructions, leaks secrets, or could be used to attack consumers)? See responsible_disclosure.md before opening a public issue.
• The LICENSE at the repo root applies to every file in the tree.