AugmentClaude

Secret Management

Read encrypted secrets at runtime in your Claude agent workflows.

Installation

  1. Make sure Claude is on your device and in your terminal.

    Skills load from ~/.claude/skills/ when Claude Code starts up — so you need it on your machine first. If you don't have it yet, install it once with the command below, then run claude in any terminal to verify.

    One-time setup
    npm i -g @anthropic-ai/claude-code

    Already have it? Skip ahead.

  2. Paste into Claude Code or into your terminal.

    This copies the whole skill folder into ~/.claude/skills/secret-yaoapp/ — the SKILL.md plus any scripts, reference docs, or templates the skill ships with. Safe default: works for every skill.

    Faster alternative (instruction-only skills)

    Skips the clone and grabs only the SKILL.md file. Don't use this if the skill ships Python scripts, reference markdowns, or asset templates — they won't be downloaded and the skill will fail when it tries to load them.

    Quick install (SKILL.md only)
    Sign up to copy
  3. Restart Claude Code.

    Quit and reopen Claude Code (or any other agent that loads from ~/.claude/skills/). New skills are picked up on startup.

  4. Just ask Claude.

    Skills auto-activate when your request matches the skill's description — no slash command needed. Trigger phrases live in the skill's own frontmatter; you can read them in the “What this skill does” section above.

Prefer to read the source first? Open on GitHub.

When Claude uses it

Claude activates this skill based on the context of your request.

What this skill does

Secret Management

Agents can read user-configured secrets at runtime using the tai tool CLI. Secrets are encrypted at rest (AES-256-GCM) and decrypted only when read.

Available Tools

ToolDescription
secret_listList secret names and descriptions (no values)
secret_readRead a single secret value by name

Usage

Bash

# List available secrets
tai tool secret_list

# Read a secret
TOKEN=$(tai tool secret_read '{"name": "GITHUB_TOKEN"}' | jq -r '.value')
git clone "https://${TOKEN}@github.com/org/repo.git"

Node.js

const { execSync } = require("child_process");

function readSecret(name) {
  const raw = execSync(
    `tai tool secret_read '${JSON.stringify({ name })}'`,
    { encoding: "utf-8" }
  );
  return JSON.parse(raw).value;
}

const token = readSecret("GITHUB_TOKEN");

Python

import json
import subprocess

def read_secret(name: str) -> str:
    result = subprocess.run(
        ["tai", "tool", "secret_read", json.dumps({"name": name})],
        capture_output=True, text=True, check=True,
    )
    return json.loads(result.stdout)["value"]

token = read_secret("GITHUB_TOKEN")

PowerShell

function Read-Secret {
    param([string]$Name)
    $json = @{ name = $Name } | ConvertTo-Json -Compress
    $result = tai tool secret_read $json | ConvertFrom-Json
    return $result.value
}

$token = Read-Secret -Name "GITHUB_TOKEN"

Security Rules

  1. Never print or log secret values — Do not write secrets to stdout, stderr, or any log file.
  2. Never write secrets to files — Exception: SSH keys may be written to ~/.ssh/ with chmod 600 permissions.
  3. Never send secrets to the LLM — Secret values must not appear in prompt content, system messages, or tool call results that are forwarded to the model.
  4. Scope isolation — Secrets are scoped per user per agent. An agent can only access secrets configured for it.
  5. Audit trail — Every secret_read call is logged in the audit trail with the caller's identity.

Related skills