DEV Community

FusionAL profile picture

FusionAL

AI tools, agent workflows, and MCP infrastructure. Building practical automation with clear permissions and oversight.

Location Williamsburg, KY Joined Joined on  Email address jrm@fusional.dev Personal website https://fusional.dev
One inbox, one kind of request, one tracker: a sensible size for a first automation

One inbox, one kind of request, one tracker: a sensible size for a first automation

Comments
2 min read

Want to connect with FusionAL?

Create an account to connect with FusionAL. You can also sign in below to proceed if you already have an account.

Already have an account? Sign in
A draft is a suggestion. A sent reply is a commitment.

A draft is a suggestion. A sent reply is a commitment.

Comments
2 min read
Write the process down first: a ten minute walkthrough decides if AI fits

Write the process down first: a ten minute walkthrough decides if AI fits

Comments
2 min read
Estimates that go quiet need a follow-up, and a draft is enough

Estimates that go quiet need a follow-up, and a draft is enough

Comments
2 min read
A voicemail becomes a work order draft, and a person checks it in under a minute

A voicemail becomes a work order draft, and a person checks it in under a minute

Comments
2 min read
The inquiry that sat in a shared inbox over the weekend

The inquiry that sat in a shared inbox over the weekend

Comments
2 min read
An unauthenticated MCP server on an internal network is still an open door

An unauthenticated MCP server on an internal network is still an open door

Comments
2 min read
A staging task that reached production started with a token in the wrong file

A staging task that reached production started with a token in the wrong file

Comments
2 min read
Secrets that never hit a git commit still reach your AI tools

Secrets that never hit a git commit still reach your AI tools

Comments
2 min read
Fewer tools, smaller blast radius: trimming an MCP server pays off twice

Fewer tools, smaller blast radius: trimming an MCP server pays off twice

Comments 1
2 min read
A log file is input too

A log file is input too

Comments
2 min read
A permissions review is only useful if someone can act on it

A permissions review is only useful if someone can act on it

1
Comments
2 min read
Evidence over assurances: what a useful AI security questionnaire answer looks like

Evidence over assurances: what a useful AI security questionnaire answer looks like

Comments
2 min read
A one-page checklist for testing one AI workflow before handoff

A one-page checklist for testing one AI workflow before handoff

Comments
2 min read
The person who sets up the integration should not own its permissions

The person who sets up the integration should not own its permissions

Comments
2 min read
A log that says "service account" does not say who did it

A log that says "service account" does not say who did it

Comments 1
2 min read
An agent that can call any URL can send your data anywhere

An agent that can call any URL can send your data anywhere

Comments 2
2 min read
The incident you have to explain is the real cost of an over-permissioned agent

The incident you have to explain is the real cost of an over-permissioned agent

Comments 1
2 min read
"It only has read access" is where the review starts

"It only has read access" is where the review starts

Comments
2 min read
Where retrieval assistants leak: the title, the summary and the cache

Where retrieval assistants leak: the title, the summary and the cache

2
Comments 3
2 min read
A summarize request should not be able to send an email

A summarize request should not be able to send an email

Comments 4
2 min read
Your AI agent is signing its work with someone else's name

Your AI agent is signing its work with someone else's name

Comments 2
2 min read
Ethical walls don't leak through the file browser. They leak through the summary.

Ethical walls don't leak through the file browser. They leak through the summary.

Comments
2 min read
loading...