A certified red teamer. A published researcher. A ghost.
For six months I published red team research on X.
Adversarial simulation frameworks....
For further actions, you may consider blocking this person and/or reporting abuse
The painful part is that publishing into the void can look like a content problem when it is often a distribution and feedback-loop problem. Technical credibility is not enough by itself. The channel has to give you some way to learn what resonated, otherwise you keep improving in private with no signal.
Alex,
This is exactly it. And I didn't realize how deep the damage went until I escaped.
For six months I was trapped in that loop: write, publish, silence. No likes, no replies, no metrics I could trust. The algorithm didn't just hide my posts - it hid the feedback itself. I couldn't tell if my content was bad, my timing was off, or if I was screaming into a sealed room. So I kept iterating in the dark, assuming the problem was me.
"Technical credibility is not enough by itself."
Hard lesson. I built frameworks, published benchmarks, validated everything. And I genuinely believed that if the work was solid, the channel would carry it. Turns out the channel was broken, not the work.
Bluesky fixed the channel. First post here - immediate signal. Likes I could see. Replies from real humans. Within 24 hours I knew more about what resonated than in six months on X. That's not because I'm suddenly better. It's because the feedback loop exists again.
Your terminalskills.io - curious if you faced similar distribution challenges building for developers? CLI tools and security research share the same problem: deeply technical audience, hard to reach, easy to lose in algorithmic noise.
Thanks for the sharp analysis. This belongs in the article itself.
That loop is brutal because it trains you to optimize for signals the platform barely gives back. The healthier metric is often not views, but whether the writing creates reusable assets: clearer positioning, sharper examples, better replies, a stronger archive. Still painful, but less dependent on the feed being generous.
Alex,
"The healthier metric is often not views, but whether the writing creates reusable assets."
This reframes everything. I've been measuring the wrong thing ā not just on X, but in my head.
Views are a platform metric. They depend on the feed being generous, the algorithm being fair, the VPN not being flagged. Reusable assets ā clearer positioning, sharper examples, a stronger archive ā those are mine. They survive platform death. They compound even when the channel is silent.
The brutal part is that X trained me to chase the wrong metric. Six months of optimizing for visibility that never came, while the actual asset ā the research itself, the frameworks, the methodology ā sat in GitHub, reusable but invisible even to me because I was too busy checking impressions.
Bluesky gives me views and the space to build assets. First post here already sharpened my positioning: "red teamer who builds benchmarks, not just breaks things." That clarity didn't come from X. It came from a channel that actually reflects back.
Your terminalskills.io ā is that the asset-first approach? CLI curation as reusable knowledge, not content-for-feed?
Yes, that is basically the idea. Terminal Skills is asset-first: the post may explain the skill, but the durable object is the reusable operating procedure.
A good skill is not content for a feed; it is a small piece of executable knowledge that can be reused by agents, humans, and future docs.
Alex,
"A good skill is not content for a feed; it is a small piece of executable knowledge that can be reused by agents, humans, and future docs."
This is the exact frame I needed. I've been treating my red team frameworks as "projectsā - things you ship and maintain. But they're not. They're skills made portable: a benchmark that teaches a methodology, a PoC that encodes a technique, a write-up that preserves a decision tree.
The redteam-ai-benchmark isn't content. It's executable knowledge. Someone can run it, extend it, or feed it to an agent that learns from it. That's the durable object. The DEV article explaining it - that's the searchable explanation. The Bluesky thread - that's the hook test and peer finder.
You just gave me vocabulary for what I've been building without naming it. Thank you.
Hey brother,
Thank you for this.
Honestly, it's one of the most thought-provoking pieces I've read in a long while.
To the certified red teamer, the published researcher,the mysterious figure behind these words: when the time is right, let this reach you.
Lately I've been seeing security come up more and more,in communities and inside companies.
Yet it's nearly always after something has already broken.Same as the law: we move only once the damage is done.People really don't change, do they?
I've taken a few "freeze beams" to my own X account as well, so this one struck rather close to home.
Perhaps people like you and me, and the rest of the security crowd, were simply a touch too stimulating for X's taste.Some signals, it seems, the algorithm would rather not read.
I'm genuinely glad our paths crossed.Finding you here on DEV, and getting to communicate even through the occasional reaction, is something I count myself lucky for. Even brushing past one another in this space has sharpened my career and raised the resolution at which I see things.
Knowing there's someone kind, and quietly proud enough,to keep writing words like these means more to me than I can easily put down. It rekindles something stubborn in me, the part that refuses to break. So let me say it plainly, even if it's a little embarrassing: a great deal of gratitude, of respect, and of love. From here.
For me it was less about leaving than about drawing a line. In my country, X leans heavily towards being a space for conversation, so I decided to lean into precisely that and keep it for communication. The denser material, the things that need links and depth, I bring here, or elsewhere. A division of labour that finally let me make my peace with the platform.
Being mutuals with you here has brought me no end of good. Thank you for that, brother.
P.S. Engineers in security burn out at a frightening rate, you know. So let's look after ourselves, and each other, and carry on living well, you and I both.
And give my regards to the ghost, the one who went and lived the part for real. Welcome back to the visible world, brother. lol
Akari,
Thank you for this. Seriously.
I wrote that article thinking I was documenting a technical failure - a platform's algorithm misfiring. But reading your words, I realized I was documenting something else entirely: loneliness. The loneliness of producing work you believe in, sending it into what you think is a room full of peers, and slowly discovering you've been talking to a mirror for six months.
Your "freeze beams" - I'm stealing that. It's perfect. Because that's exactly what it feels like: not a ban, not a punishment, just a cold, silent pause that never ends. And the worst part is you don't even feel the cold until someone else points out you've been frozen.
"Even brushing past one another in this space has sharpened my career."
This hit me hard. Because while I was a ghost, I had no idea I was brushing past anyone. I thought I was alone in the dark. Knowing that my work reached you even when the platform told me it reached no one - that rewrites the whole story. It doesn't make the ban less real, but it makes the invisibility less absolute. Thank you for telling me this.
Your division of labour is wise. Keeping X for conversation, bringing the dense material to places that respect it - that's not compromise, that's strategy. I might adopt something similar, though honestly, after the ghosting, I'm reluctant to give X even that much of my time. But your approach is pragmatic, and I respect pragmatism in our field. We need more of it.
"Engineers in security burn out at a frightening rate."
Yes. We do. And we pretend we don't. We talk about resilience and grind and "the mission," but we rarely admit that the mission doesn't care if we break. So your proposal - to look after ourselves and each other - I'll take that seriously. Consider this my acceptance of the pact. If you ever feel the freeze closing in, or the weight of the work getting too loud, you have someone here who understands exactly what that silence feels like.
And thank you for welcoming me back to the visible world. I didn't expect to find kindred spirits so quickly after escaping the sandbox. But here you are, proving that the communities we actually need are the ones built on protocols, not on one man's algorithm.
Let's keep each other visible, Akari. And let's keep each other human.
With respect, gratitude, and solidarity -
toxy4ny
I can relate lol. I had a similar problem with my github. Honestly shadowbanning should never have existed.
Mirrai,
Same here. And yeah - GitHub, X, doesn't matter which platform. The mechanism is identical: algorithm decides you're noise, and you're gone. No appeal, no explanation, just silence.
"Shadowbanning should never have existed."
Hard agree. It's not moderation. It's cowardice. A platform that can't look you in the eye and say "we're limiting you" has no right to host communities.
Your profile says you break things to understand them. Same. And it's ironic - we build tools to expose systems, but the systems that host us refuse to be exposed. No CVE for "platform silently ghosts its own users."
Good to know you're here. Let's keep breaking things - and documenting what breaks us.