hydra-head

• Self-calibrating: Skills detect your installed CLI version and adapt commands automatically
• Safe by design: Risky operations require explicit human invocation
• Least privilege: Each skill has restricted tool access appropriate to its function
• Token-efficient: Compact frontmatter with progressive detail loading

Option A: Vercel Skills CLI

```bash

npx skills add Flux-Point-Studios/cardano-agent-skills

```

Option B: add-skill (supports skill selection)

```bash

# Install all skills

npx add-skill Flux-Point-Studios/cardano-agent-skills -a claude-code

# Install specific skills only

npx add-skill Flux-Point-Studios/cardano-agent-skills --skill cardano-cli-wallets --skill cardano-cli-transactions -a claude-code

```

Core CLI Operations

| Skill | Description | Risk Level |

|-------|-------------|------------|

| cardano-cli-doctor | Diagnose CLI version, detect era-prefixed vs legacy syntax, produce compatibility report | Safe (read-only) |

| cardano-cli-wallets | Create/manage keys, addresses, UTxO checks, wallet dossier output | Safe (guidance) |

| cardano-cli-wallets-operator | Execute wallet operations (key generation, address building) | Manual invoke |

| cardano-cli-transactions | Build, sign, submit standard transactions (guidance + templates) | Safe (guidance) |

| cardano-cli-transactions-operator | Execute transaction builds and submits | Manual invoke |

| cardano-cli-staking | Stake key registration, delegation, rewards withdrawal (guidance) | Safe (guidance) |

| cardano-cli-staking-operator | Execute staking operations | Manual invoke |

| cardano-cli-plutus-scripts | Plutus script transactions: datums, redeemers, collateral (guidance) | Safe (guidance) |

| cardano-cli-plutus-scripts-operator | Execute script spends and submits | Manual invoke |

| cardano-protocol-params | Fetch and validate protocol parameters | Safe |

Smart Contracts

| Skill | Description | Risk Level |

|-------|-------------|------------|

| aiken-smart-contracts | Aiken workflows: validators, building, blueprints, .plutus generation | Safe |

| plutus-v3-conway | Plutus V3 under Conway: contexts, governance, V2→V3 migration | Safe |

| meshjs-cardano | MeshJS patterns: tx building, UTxO selection, wallet connectors | Safe |

Hydra L2

| Skill | Description | Risk Level |

|-------|-------------|------------|

| hydra-head | Hydra Head best practices: setup, keys, peers, lifecycle (guidance) | Safe (guidance) |

| hydra-head-operator | Execute Hydra operations (init, commit, close) | Manual invoke |

| hydra-head-troubleshooter | Decision tree for Hydra issues: symptoms → fixes → verification | Safe |

```

cardano-agent-skills/

├── shared/

│ └── PRINCIPLES.md # Common safety rules across all skills

├── skills/

│ ├── /

│ │ ├── SKILL.md # Skill definition (frontmatter + instructions)

│ │ ├── reference/ # Deep-dive docs, patterns, examples

│ │ ├── templates/ # Copy-paste templates, worksheets

│ │ └── examples/ # Expected output samples

│ └── ...

├── .github/

│ └── workflows/

│ └── validate-skills.yml # CI validation

└── README.md

```

1. Self-Calibrating (Dynamic Context)

Skills that interact with CLI tools use dynamic context injection to adapt to your installed version:

```yaml

---

name: cardano-cli-doctor

context:

- "!cardano-cli version"

- "!cardano-cli --help | head -40"

- "!cardano-cli conway --help 2>&1 | head -20"

---

```