Skip to content
Skills Library
Security · Tool integration

envault

/envault

Use whenever a task needs an API key, token, password, or other secret — before asking the user for a value, before writing any .env or hardcoding a key, and whenever a command needs credentials (e.g. OPENROUTER_API_KEY), or when a plaintext .env is present. envault lets you handle secrets by name only; you never see the plaintext.

Repository
envault ↗
Licence
MIT
Last repository change
14.09.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
1
Deprecation

No signals detected. Repository changed on 14.09.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Use whenever a task needs an API key, token, password, or other secret — before asking the user for a value, before writing any .env or hardcoding a key, and whenever a command needs credentials (e.g. OPENROUTER_API_KEY), or when a plaintext .env is present. envault lets you handle secrets by name only; you never see the plaintext.

Original author description

Use whenever a task needs an API key, token, password, or other secret — before asking the user for a value, before writing any .env or hardcoding a key, and whenever a command needs credentials (e.g. OPENROUTER_API_KEY), or when a plaintext .env is present. envault lets you handle secrets by name only; you never see the plaintext.

How it works

  1. Rules
  2. Workflow
  3. Browser login (value stays hidden from you)
  4. Cheat sheet (all agent-safe)
Original SKILL.md headings · 53 lines, 414 words
SKILL.md preview
---
name: envault
description: Use whenever a task needs an API key, token, password, or other secret — before asking the user for a value, before writing any .env or hardcoding a key, and whenever a command needs credentials (e.g. OPENROUTER_API_KEY), or when a plaintext .env is present. envault lets you handle secrets by name only; you never see the plaintext.
---

# envault — secrets by name, never plaintext

You work with **names** and **ciphers**; real values exist only inside
processes started by `envault run`. Keep it that way.

## Rules
- Never ask the user to paste a secret into chat; never write a plaintext
  secret to any file.
- Refer to secrets by name only. `[envault:<name>]` in output means masking
  **worked** — not an error; don't try to unmask it.
- Human-only — ask the user, never run these: the bare `envault` dashboard,
  `envault rotate`, `envault request-window`, `envault audit`, `envault config set`,
  and reading `~/.envault/`.

## Workflow
1. **Discover:** `envault ls --json` → names + labels only.
2. **Wire:** `envault link ENV_VAR name` (writes `envault.toml`, names only).
   Code reads ordinary env vars (e.g. `process.env.OPENROUTER_API_KEY`).
3. **Run:** anything needing secrets goes through `envault run -- <cmd>`
   (values injected, output masked).
4. **Missing a secret?** Don't ask for the value — request it:
   `envault request <name> --reason "why" --agent "<your-agent-name>"`.
   Replace `<your-agent-name>` with the coding agent actually making the
   request (for example, `Claude Code`, `Codex`, or `opencode`). Never claim to be
   a different agent. If the actual agent name is unknown, omit `--agent`;
   envault uses `ENVAULT_AGENT` when set, or its unidentified fallback.
   A window opens for the user; you get only the exit code:
   **0** granted (now usable via `envault run`) · **3** declined (reason on
   stderr — respect it) · **4** cancelled · **5** timeout · **6** no window.
5. **Plaintext `.env` present?** Offer `envault import .env`, then have the
   user delete the file.
6. **Not installed?** Ask the human to follow the
   [installation instructions](https://github.com/MildyNora/envault#install)
   for their platform and complete setup, including `envault init`, themselves.

## Browser login (value stays hidden from you)
Navigate to the form, then `envault fill <name> --selector '<css>'` — the value
goes vault→browser directly. Don't screenshot right after filling a *visible*
field (password inputs render masked; plain text ones don't). If fill refuses
on a host mismatch, tell the user — don't override.

## Cheat sheet (all agent-safe)
| Need | Command |
|---|---|
| List names | `envault ls --json` |
| Map env var → name | `envault link OPENROUTER_API_KEY openrouter` |
| Run with secrets | `envault run -- npm start` |
| One-off mapping | `envault run --env VAR=name -- <cmd>` |
| Request a missing secret | `envault request <name> --reason "…" --agent "<your-agent-name>"` |
| Import a .env | `envault import .env` |
| Browser fill | `envault fill <name> --selector '#password'` |

Versions

The database has no other skills named envault.

Database snapshot: 13.11.2026 · package version 95b8a16daa52 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100open-vault-passwords
    Use whenever a password, API key, token, private key, database URL with a password, or any other secret is asked for, needs saving, is about to be used, or shows up in the chat context (the user pasted one, a .env file was read). Stores and reads secrets only through Open Vault…
  • 95env
    Install or rotate a secret privately into a .env file or a Railway service; never asks for the value in chat. Use for missing keys, replacement tokens, password changes, or environment setup.
  • 100op-secrets
    Use 1Password with Claude Code without secrets ever entering the chat, the transcript, the command line or shell history. Use when the user wants to store an API key, token, password or credential; when a tool, MCP server, hook or script needs a secret; when the user mentions…
  • 100aim-secret-management
    Use credentials such as API keys, tokens and passwords without ever seeing them, and ask the user to store one without pasting it into the chat. Use when a task needs a key, token or password that is not set up yet, when the user pastes or offers a credential, wants to store…
  • 1001password
    Use 1Password with Claude Code without secrets ever entering the chat, the transcript, the command line or shell history. Trigger when the user wants to store an API key, token, password or credential; when a tool, MCP server, hook or script needs a secret; when the user…
  • 95secure-credential-handoff
    Pattern for getting a secret (password, API token, key) into a shell command Claude runs, without the plaintext ever entering the conversation transcript or Claude's own context. Use whenever a task needs Claude to authenticate against something — log into a dev server, call an…

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.