Share secrets at lightning speed

One-time secret sharing API. Create a secret, share the link, it self-destructs after viewing. We log events — created, viewed, expired and destroyed — but the secret's content is never logged.

No account needed. 1 view, expires in 1 hour, up to 10 KB.

How It Works

πŸ”‘

1. Create a Secret

Use the form above, the dashboard or the API to create a secret with content, an optional passphrase and an expiration.

πŸ”—

2. Share the Link

Send the one-time link to anyone. Share the passphrase through a different channel for extra security.

πŸ”₯

3. It Self-Destructs

After its last allowed view, or when it expires, the link stops working instantly and the encrypted content is erased (an expired secret within 10 minutes). The rest of the record is deleted from our database 30 days later (for backups, see Security).

What People Use It For

Send a .env to a teammate

Paste the file into the dashboard or send it through the API, allow a single view, and share the passphrase over a different channel. Once your teammate opens the link, it stops working for everyone else. How to share a .env file.

Send a password to a client

Your client needs no account — just the link. Set an expiry so an unopened link doesn't linger in their inbox, and add a passphrase you tell them on the phone. How to share a password with a client.

Hand a secret to CI/CD

Give the pipeline a read-only API key and fetch the secret in a build step (see the safe export example below). Every fetch uses up a view, so this suits one-off handoffs — bootstrapping a new environment — not values a pipeline reads on every run. In a pipeline, drop the cat line: it prints the values into the build log.

Ship a .env to Docker

Fetch the secret once as an env file for docker compose or docker stack deploy — or turn every key into a swarm secret in one go. Values with $, quotes or # arrive exactly as written. One view per fetch, so it fits setting up a server, not every deploy. See the Docker docs.

Built for Developers

Create

curl -X POST \
  https://secretbolt.io/api/v1/secrets \
  -H "Authorization: Bearer $KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Production DB",
    "content": "DATABASE_URL=...",
    "max_views": 1,
    "expires_in_hours": 24
  }'

Retrieve

# One-time retrieval
curl https://secretbolt.io/api/v1/secrets/$TOKEN \
  -H "Authorization: Bearer $KEY" \
  -H "X-Passphrase: $PASSPHRASE"

# Recommended: write to a private file, read it, then load it
(umask 077 && curl -fsS -o .env.secretbolt \
  -H "Authorization: Bearer $READ_ONLY_KEY" \
  https://secretbolt.io/api/v1/secrets/$TOKEN/export \
  -H "X-Passphrase: $PASSPHRASE")   # leave $PASSPHRASE empty if the secret has none
cat .env.secretbolt          # check what you're about to load
set -a; . ./.env.secretbolt; set +a
rm .env.secretbolt

Shortcut: eval

The output is escaped and PATH or LD_PRELOAD are never exported, but eval means trusting every server response without looking at it.

secret_env=$(curl -fsS \
  -H "Authorization: Bearer $READ_ONLY_KEY" \
  https://secretbolt.io/api/v1/secrets/$TOKEN/export \
  -H "X-Passphrase: $PASSPHRASE") &&   # leave $PASSPHRASE empty if the secret has none
  eval "$secret_env"

Each fetch uses up one view of the secret, so fetch once and keep the file until its values are loaded where you need them.

Security

What actually happens to a secret, and what we can and can't see.

Encrypted before it is stored

Content is encrypted with AES-256-GCM (Rails Active Record Encryption), with a fresh random IV per secret, before it is written to the database. The database never holds plaintext.

Keys live outside the database

The encryption keys sit in the app's encrypted credentials, and the key that unlocks those is handed to the server at deploy time. A copy of the database alone cannot be decrypted.

Server-side, not end-to-end

Encryption happens on our server, which decrypts the content to show it to the recipient. This is not end-to-end encryption: while serving a secret, the server handles its plaintext.

Passphrases and links

A passphrase is stored only as a bcrypt hash. Every link carries a random 192-bit token, so links cannot be guessed.

Loading a secret into your shell

The /export endpoint single-quotes every value and never exports variables that run code, such as PATH or LD_PRELOAD. Even so, we recommend saving it to a file only you can read, reading that file, and only then loading it. Piping it straight into eval is the quicker shortcut, at the cost of trusting the response without seeing it.

What gets deleted, and when

A secret stops being served the moment its last allowed view is used or it expires. At that moment its encrypted content and passphrase hash are erased from the database (an expired secret within 10 minutes). The rest of its record and its event log are permanently deleted 30 days later. Encrypted daily database backups are kept for 30 days, so erased content can remain in a backup for up to 30 days.

Secrets created without an account

A secret made with the form at the top of this page is encrypted, logged and deleted exactly like any other. It is tied to no account, and your IP address is not stored with it — it is only used to count the limit of 5 secrets an hour. It can be opened 1 time and expires after 1 hour.

What we log

Each secret's events: created, viewed, expired and destroyed — with the viewer's IP address and user agent for views. The event log never records the content, the passphrase or the link.

Ready to stop leaking secrets?

Free to start. No credit card required.

Get Started Free