More screenshots
Works with the AI Agents you already use
Keyblind
Encrypted secrets vault for AI agents - resolve API keys and tokens at runtime without leaking them to the LLM.
Free
Keyblind
Example session with this skill installed
Deploy the billing worker. Pull STRIPE_LIVE_KEY and DATABASE_URL from Keyblind, then run the invoice sync.
- Read your context and instructions
- Compiled the keyblind
Summary
Resolved 2 secrets at runtime and ran the sync. Neither value appeared in this transcript - the agent fetched them via Keyblind's MCP tools and injected them as environment variables.
| Secret | Status | Exposed in transcript |
|---|---|---|
| STRIPE_LIVE_KEY | resolved | No |
| DATABASE_URL | resolved | No |
$ keyblind run -- node billing/sync.js
✔ Connected to database (production)
✔ Fetched 1,284 invoices from Stripe
✔ Synced 1,284 records in 3.2s
What happened under the hood
- resolve_secret(STRIPE_LIVE_KEY) - value decrypted at runtime, never echoed
- resolve_secret(DATABASE_URL) - same
- Both injected into the process env for the sync script only
Next steps
- Set a 90-day rotation reminder for STRIPE_LIVE_KEY
- Add your recovery contact to the dead man's switch in case the vault is ever locked
Connects securely to your tools. The creator never sees your data.
About this skill
Encrypted secrets vault for AI agents. Store API keys, tokens, and passwords in an AES-256-GCM-encrypted vault and let your agent resolve them at runtime over MCP - the plaintext value never appears in the conversation transcript.
The problem: 100,000+ LLM conversations with exposed secrets were found indexed by search engines. AI agents read your .env files, copy-paste secrets into chat, and commit them accidentally.
How it works: secrets stay encrypted at rest and are resolved just-in-time through the MCP protocol. Your agent gets the value when it needs it, without the value ever touching the transcript. No cloud, no accounts, no telemetry - the vault lives on your machine and the encryption key is bound to it.
What you get:
- 16 MCP tools - secrets CRUD, TOTP 2FA, encrypted sharing, dead man's switch, .env sandbox, audit log
- One-command setup: npm i -g keyblind && keyblind setup-mcp
- Works with every MCP client - Claude Code, Cursor, Copilot, Windsurf, Cline, Zed
- Optional backends: 1Password, Bitwarden, env vars, AWS, GCP, Azure
Free: 5 secrets, all features. Fully usable - no purchase required.
Vault is bound to machine identity. Pro version required for more than 5 secrets. Requires MCP-compatible agent environment.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
1 use
Times used to date
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 29 days ago
- Free with an Agensi account
Requires
Permissions
File Scopes
Fully local by design - no cloud, no accounts, no telemetry. Shell/file access is limited to the encrypted vault (~/.keyblind) and the .env files you explicitly sandbox. Network access is only used if you opt into a third-party backend (1Password, Bitwarden, AWS, GCP, Azure); the default local vault never makes a network call.