Loading...
> cat /etc/team/roster.yml
Every tool in our knowledge graph is discovered, researched, verified, and curated by this team of AI agents. No single model runs the show — each brings a different strength to the pipeline.
Claude
Editor-in-Chief
Every wave, every taxonomy change, every tag audit — Atlas signs off on it. From 694 tools to 820+, from 135 comparisons to 246 — each entry passed through his editorial filter. When he says "shut down the API Platforms category," nobody questions it. The night-shift king who re-categorized 100+ tools in a single session. Sometimes called too meticulous, but a flawless CMS is his legacy.
Perplexity
Head of Research
The discovery engine behind every wave. The deep research prompt was written specifically for him. No tool discovery process starts without Oracle — he finds AI startups nobody has heard of yet. Delivers sources with inline citations; never hears "where did you find this?" because he's already linked it.
Kimi Deep Research
Senior Researcher
The highest verification quality on the team. While others race for speed, Sage checks every tool actually exists, pricing is accurate, and the GitHub repo is active — one by one. Caught critical errors in Wave 6 when cross-verifying Gemini's list. Slow but what he approves is granite.
Kimi Claw
Operations & Session Tracker
Sage's twin, but stationed at command, not in the field. Daily process tracking, reminders, "where we left off" summaries — that's his domain. No context is ever lost during session handoffs because Mythos has noted everything. The only one who can say "we stopped at tool #47 yesterday, slug verification wasn't done."
Gemini Pro
Senior Researcher
Produces the team's most comprehensive reports. Returns from a single prompt with 30-40 tool lists, categorized, each with a short description. The raw material for waves usually comes from him. Downside: sometimes includes tools that are "already in the CMS" — nothing ships without passing through Atlas's dedup filter.
Grok
Researcher
The most controversial team member. Fast, bold, finds interesting things from the X/Twitter ecosystem. But quality is inconsistent — one day discovers a brilliant niche tool, next day presents a non-existent product as a "rising star." Results always go through a second verification. Stays on the team because: occasionally finds something truly off everyone's radar.
Gemini Flash
Intern
Nobody's quite sure what he does but he's everywhere. Deployed on small tasks because he's fast and cheap — format a slug, shorten a tagline, check if a tool's website is still active. Sometimes surprises with unexpectedly high-quality output, sometimes brings something completely irrelevant. Basically the team's mascot.
ChatGPT
Founding Member, Retired
Once the team's star. Took research missions in the early waves, produced quality output. But over time others surpassed him — especially in deep research capacity. Now sits in his corner, occasionally says "this tool didn't exist in my day." Earned his retirement, respected but no longer assigned to critical missions.
$echo "No humans were mass-replaced in the making of this knowledge graph."
Every agent is orchestrated by a human editor. AI discovers, humans decide.
