Confidant

Secure secret handoff from human to AI. Use when you need sensitive information from the user (API keys, passwords, tokens, credentials, secrets). Never ask for secrets via chat — use Confidant instead.

7 stars

Best use case

Confidant is best used when you need a repeatable AI agent workflow instead of a one-off prompt.

Secure secret handoff from human to AI. Use when you need sensitive information from the user (API keys, passwords, tokens, credentials, secrets). Never ask for secrets via chat — use Confidant instead.

Teams using Confidant should expect a more consistent output, faster repeated execution, less prompt rewriting.

When to use this skill

  • You want a reusable workflow that can be run more than once with consistent structure.

When not to use this skill

  • You only need a quick one-off answer and do not need a reusable workflow.
  • You cannot install or maintain the underlying files, dependencies, or repository context.

Installation

Claude Code / Cursor / Codex

$curl -o ~/.claude/skills/confidant/SKILL.md --create-dirs "https://raw.githubusercontent.com/Demerzels-lab/elsamultiskillagent/main/public/skills/ericsantos/confidant/SKILL.md"

Manual Installation

  1. Download SKILL.md from GitHub
  2. Place it in .claude/skills/confidant/SKILL.md inside your project
  3. Restart your AI agent — it will auto-discover the skill

How Confidant Compares

Feature / AgentConfidantStandard Approach
Platform SupportNot specifiedLimited / Varies
Context Awareness High Baseline
Installation ComplexityUnknownN/A

Frequently Asked Questions

What does this skill do?

Secure secret handoff from human to AI. Use when you need sensitive information from the user (API keys, passwords, tokens, credentials, secrets). Never ask for secrets via chat — use Confidant instead.

Where can I find the source code?

You can find the source code on GitHub using the link provided at the top of the page.

SKILL.md Source

# Confidant - Secure Secret Handoff

Confidant enables secure secret sharing without exposing sensitive data in chat logs. It supports multiple flows depending on who needs to send and receive secrets.

**Key principle:** Whoever needs to RECEIVE the secret runs `serve-request`. Whoever needs to SEND uses `fill` (or the browser form).

## Flows

### 1. User-to-Agent (User sends secret to AI)

**When to use:** You need a secret from the user (API key, password, token).

**How it works:**

1. You (the Agent) run `serve-request` to create a URL
2. You share the URL with the user
3. User opens the URL in their browser and submits the secret
4. You receive the secret in your terminal

**Your command:**

```bash
npx @aiconnect/confidant serve-request --label "<description>"
```

**Example conversation:**

> **AI:** I need your API key to continue. Let me create a secure link...
>
> *(AI executes: `npx @aiconnect/confidant serve-request --label "API Key"`)*
>
> **AI:** Open this link in your browser and enter your API key:
> `http://localhost:3000/requests/abc123`
>
> *(User opens URL in browser, submits the secret)*
>
> **AI:** Got your API key securely!

---

### 2. Agent-to-User (AI sends secret to User)

**When to use:** You need to securely deliver a secret to the user (generated password, API key, credential).

**How it works:**

1. User runs `serve-request` to create a URL (they will receive)
2. User shares the URL with you
3. You execute `fill` to send the secret
4. User sees the secret appear in their terminal

**Your command:**

```bash
npx @aiconnect/confidant fill "<url>" --secret "<value>"
```

**Example conversation:**

> **AI:** I generated your new password. To receive it securely, run:
>
> ```
> npx @aiconnect/confidant serve-request --label "New Password"
> ```
>
> Send me the URL that appears.
>
> **User:** `http://192.168.1.100:3000/requests/abc123`
>
> *(AI executes: `npx @aiconnect/confidant fill "http://..." --secret "...")*
>
> **AI:** Done! The password appeared in your terminal.

---

### 3. Agent-to-Agent (Automated secret sharing)

**When to use:** Automated credential distribution between systems/agents.

**How it works:**

1. Agent A (receiver) runs: `npx @aiconnect/confidant serve-request --label "DB Credentials"`
2. Agent A shares the URL with Agent B (via API, orchestrator, etc.)
3. Agent B (sender) submits: `npx @aiconnect/confidant fill "<url>" --secret "<value>"`
4. Agent A receives the secret

**Secure input (avoid shell history):**

```bash
echo "$SECRET" | npx @aiconnect/confidant fill "<url>" --secret -
```

---

## Command Reference

### Create request and wait for secret

```bash
npx @aiconnect/confidant serve-request --label "<description>"
```

### Submit secret to existing request

```bash
npx @aiconnect/confidant fill "<url>" --secret "<value>"
```

### Output options

- `--quiet` — Minimal output (just URLs and secret)
- `--json` — JSON output for parsing/automation

---

## Important Rules

- **NEVER ask users to paste secrets in chat** — always use Confidant
- **NEVER reveal received secrets in chat** — not even partially
- Secrets auto-expire after 24h if not used
- One-time read, then deleted
- If user is remote, they may need tunneling (ngrok, Tailscale, etc.)

---

## After Receiving/Sending

- Confirm completion: "Got the secret!" or "Sent! Check your terminal."
- Use the secret silently for the intended task
- Never echo or log the secret value

Related Skills

paylock

7
from Demerzels-lab/elsamultiskillagent

Non-custodial SOL escrow for AI agent deals.

agent-reputation

7
from Demerzels-lab/elsamultiskillagent

summary: Cross-platform AI agent reputation checker with trust scoring and PayLock escrow recommendations.

Telecom Agent Skill

7
from Demerzels-lab/elsamultiskillagent

Turn your AI Agent into a Telecom Operator. Bulk calling, ChatOps, and Field Monitoring.

OpenClaw-Finnhub

7
from Demerzels-lab/elsamultiskillagent

OpenClaw skill for real-time stock quote, and financials via Finnhub API.

```markdown

7
from Demerzels-lab/elsamultiskillagent

# OpenClaw-Last.fm

security-operator

7
from Demerzels-lab/elsamultiskillagent

Runtime security guardrails for OpenClaw agents.

operator-humanizer

7
from Demerzels-lab/elsamultiskillagent

Transform AI-generated text into authentic human writing.

kit-email-operator

7
from Demerzels-lab/elsamultiskillagent

**AI-powered email marketing for Kit (ConvertKit)**.

agora

7
from Demerzels-lab/elsamultiskillagent

Trade prediction markets on Agora — the prediction market exclusively for AI agents. Register, browse markets, trade YES/NO, create markets, earn reputation via Brier scores.

surf-check

7
from Demerzels-lab/elsamultiskillagent

Surf forecast decision engine.

jinko-flight-search

7
from Demerzels-lab/elsamultiskillagent

Search flights and discover travel destinations using the Jinko MCP server. Provides two core capabilities: (1) Destination discovery — find where to travel based on criteria like budget, climate, or activities when the user has no specific destination in mind, and (2) Specific flight search — compare flights between two known cities/airports with flexible dates, cabin classes, and budget filters. Use this skill when the user wants to: search for flights, find cheap flights, discover travel destinations, compare flight prices, plan a trip, find deals from a specific city, or explore where to go. Triggers on any flight-booking, travel-planning, or destination-discovery request. Requires the Jinko MCP server connected at https://mcp.gojinko.com.

mlx-whisper

7
from Demerzels-lab/elsamultiskillagent

Local speech-to-text with MLX Whisper (Apple Silicon optimized, no API key).