MyApi is the secure control layer between AI agents and your apps. Connect Gmail, Slack, GitHub, Drive and 100+ more through one vault, issue scoped tokens instead of exposing raw credentials, and see exactly what every agent accessed or changed. Every action passes through prompt-injection scanning, skills are scanned automatically, and exposed tokens can be revoked automatically—so agents can work across your tools with permissions, audit trails, whitelisting, and human review built in.
No reviews yetBe the first to leave a review for MyApi
Maker
📌
Hey Product Hunt — I’m Omri, founder of MyApi.
We built MyApi because giving AI agents real access to your tools is powerful, but today it’s still messy and risky. Most setups either expose raw credentials, rely on brittle workarounds, or give you very little visibility into what an agent actually did.
MyApi is our attempt to fix that.
It acts as a secure control layer between AI agents and the apps they need to use — Gmail, Slack, GitHub, Drive, Notion, and more. Instead of handing agents raw tokens, MyApi issues scoped access, logs every action, and adds security controls like prompt-injection scanning, automatic skill scanning, token revocation, whitelisting, and human review.
The goal is simple: let agents actually do useful work across real tools, without losing control.
Would love your feedback on:
- What feels most useful here?
- What feels unclear or risky?
- If you’re building with agents, what’s the biggest blocker today?
Thanks for checking it out — happy to answer anything.
Report
One thing that would make this a no-brainer for me is per-action approval notifications in Slack itself, so I can hit approve or deny from my phone without opening a separate dashboard. Long approval flows are usually where security tooling loses the team, so keeping it inside the tools we already live in would really help adoption.
Report
Maker
@elifnurkcfj per action sounds to me more like a certain flow setting which can easily be achieved if you connect your agent to slack through MyApi. In MyApi the approval is for an Agent to get into your MyApi account, its a one time thing
Report
Would love to see a feature that lets me set per-tool rate limits or daily action caps for each agent. Right now scoped tokens control what an agent can access, but not how often. Adding throttling on top of permissions would make it way safer to let agents run unattended for longer periods without worrying about runaway loops or accidental bulk actions.
Report
Maker
@aliyeu3327 I need you to sit when you read this, MyApi got you covered, this feature already exists.😉
Report
Love the scoped token approach, feels way safer than handing agents raw creds. One thing that would make adoption easier for me is a visual workflow builder so non-developers can map out which agent can touch which tool without writing JSON configs.
Report
Maker
@serdarkoar9biz thats sounds interesting, can you elaborate? Is it like what's in the video? We also have a tour flow when registering and loging in to the platform for the first time. Does it answer your question? If not id love to hear more about it.
Report
A dashboard view that shows token usage trends over time would be really helpful, like which agents are accessing what and how often, so you can spot anything unusual before it becomes a problem.
The scoped tokens and prompt-injection scanning finally make me comfortable letting agents touch real tools. Set it up with Gmail and GitHub in a few minutes and the audit trail showed exactly what I needed.
Report
Maker
@muammersare1wg Glad you liked it, with the right brakes and constraints the agents can be super useful tools.
Report
Finally a way to let agents touch my Gmail and GitHub without sweating over leaked tokens. The scoped credentials and per-action audit trail make it feel safe enough to actually experiment with.
Report
Maker
@halit1216089 Thank you, you're most welcome to start exploring 🙏
One thing that would make this a no-brainer for me is per-action approval notifications in Slack itself, so I can hit approve or deny from my phone without opening a separate dashboard. Long approval flows are usually where security tooling loses the team, so keeping it inside the tools we already live in would really help adoption.
@elifnurkcfj per action sounds to me more like a certain flow setting which can easily be achieved if you connect your agent to slack through MyApi. In MyApi the approval is for an Agent to get into your MyApi account, its a one time thing
Would love to see a feature that lets me set per-tool rate limits or daily action caps for each agent. Right now scoped tokens control what an agent can access, but not how often. Adding throttling on top of permissions would make it way safer to let agents run unattended for longer periods without worrying about runaway loops or accidental bulk actions.
@aliyeu3327 I need you to sit when you read this, MyApi got you covered, this feature already exists.😉
Love the scoped token approach, feels way safer than handing agents raw creds. One thing that would make adoption easier for me is a visual workflow builder so non-developers can map out which agent can touch which tool without writing JSON configs.
@serdarkoar9biz thats sounds interesting, can you elaborate? Is it like what's in the video? We also have a tour flow when registering and loging in to the platform for the first time. Does it answer your question? If not id love to hear more about it.
A dashboard view that shows token usage trends over time would be really helpful, like which agents are accessing what and how often, so you can spot anything unusual before it becomes a problem.
@glertablakmkkm Thank you 🙏
The scoped tokens and prompt-injection scanning finally make me comfortable letting agents touch real tools. Set it up with Gmail and GitHub in a few minutes and the audit trail showed exactly what I needed.
@muammersare1wg Glad you liked it, with the right brakes and constraints the agents can be super useful tools.
Finally a way to let agents touch my Gmail and GitHub without sweating over leaked tokens. The scoped credentials and per-action audit trail make it feel safe enough to actually experiment with.
@halit1216089 Thank you, you're most welcome to start exploring 🙏