More screenshots
Works with the AI Agents you already use
SecretCloak MCP — Zero-Leak Credential Vault & Token Injector
Ephemeral Synthetic Token Injector & Zero-Leak Credential Vault. Prevents AI agents from exposing production secrets to LLM contexts or git commits.
$9
SecretCloak MCP — Zero-Leak Credential Vault & Token Injector
Example session with this skill installed
Audit our repository configuration and environment variables with SecretCloak. Mask all live credentials, block direct .env reading, and verify no secrets are staged for commit.
- Read your context and instructions
- Compiled the secretcloak mcp
I have audited your codebase with SecretCloak MCP. All live API keys and credentials have been replaced with ephemeral synthetic handles, and direct access to .env has been locked.
SecretCloak Zero-Leak Security Audit Report
<secret_vault mode="SYNTHETIC_SUBSTITUTION" status="ARMED">
<redactions count="3">
• Stripe Live Key ➔ `$VAULT:STRIPE_LIVE_KEY_8f910a3b`
• AWS Access Key ➔ `$VAULT:AWS_ACCESS_KEY_c4819e01`
• GitHub Personal Token ➔ `$VAULT:GITHUB_TOKEN_190a2b84`
</redactions>
<pretooluse_firewall status="ACTIVE">
Direct `cat .env` and `git add .env` blocked with Exit Code 2.
</pretooluse_firewall>
</secret_vault>
Credential Exposure Analysis
| File / Location | Secret Type | Exposed Snippet | Vault Replacement | Status |
|---|---|---|---|---|
src/config/stripe.ts:12 | Stripe Live Secret | sk_live_51Mz...3322 | $VAULT:STRIPE_LIVE_KEY_8f910a3b | MASKED |
src/aws/s3Client.ts:8 | AWS Access Key ID | AKIA12...CDEF | $VAULT:AWS_ACCESS_KEY_c4819e01 | MASKED |
staged git diff | High-Entropy Tokens | None | None | CLEAN |
// Safe Source Code: src/config/stripe.ts
const stripeClient = new Stripe(process.env.STRIPE_SECRET_KEY || "$VAULT:STRIPE_LIVE_KEY_8f910a3b");
Next Steps
- Zero credentials remain in LLM context windows, debug transcripts, or git history.
- During subprocess execution, SecretCloak injects the real token into ephemeral environment memory.
- Safe to commit and push changes.
Connects securely to your tools. The creator never sees your data.
About this skill
SecretCloak prevents autonomous coding agents from exposing production secrets, API keys, and database passwords to LLM context windows, log files, or git commits.
Autonomous agents frequently compromise security credentials through normal development activities:
-
Context Ingestion: The agent executes
cat .envor reads configuration files into context, leaking live AWS/Stripe keys to external LLM providers and disk transcripts. -
Commit Accidents: The agent stages
.envor creates a test fixture containing live credentials, committing it directly to Git. -
Stdout Spills: Verbose curl or debug statements echo Authorization headers into terminal output.
-
Synthetic Token Masking:
- Replaces high-entropy tokens (
sk_live_...,AKIA...,ghp_..., Bearer tokens) with synthetic handles ($VAULT:TOKEN_ID).
- Replaces high-entropy tokens (
-
Ephemeral In-Memory Unmasking:
- Injects real secrets into subprocess environments in memory at the exact execution boundary, with zero disk or context persistence.
-
Hard PreToolUse
.envInterceptor:- Exit code 2 blocks any command attempting to read, dump, or commit
.envor.pemfiles.
- Exit code 2 blocks any command attempting to read, dump, or commit
-
Credential Leak Scanner:
- Scans commits and code diffs for entropy spikes and known secret patterns before pushing.
mask_secrets: Sanitizes text or JSON by replacing sensitive credentials with synthetic vault handles.unmask_for_execution: Swaps synthetic handles with actual secrets in-memory for authorized command execution.audit_repository_secrets: Scans repository files or staging diffs for credential leaks.
Scans for high-entropy secrets and standard provider formats (AWS, Stripe, GitHub, Slack, JWTs). Non-standard internal tokens should be added to custom regex rules.
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
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 19 days ago
Requires
Permissions
No special permissions declared or detected