afro (@afrow0w13) header
afro (@afrow0w13) profile picture

Building in Web2 & Web3 | Architecting software | AI & LLM Enthusiast | Living crypto

Joined September 2014
Photo by afro (@afrow0w13)Photo by afro (@afrow0w13)Photo by afro (@afrow0w13)Photo by afro (@afrow0w13)Photo by afro (@afrow0w13)Photo by afro (@afrow0w13)Photo by afro (@afrow0w13)Photo by afro (@afrow0w13)Photo by afro (@afrow0w13)Photo by afro (@afrow0w13)
I built RunWeaver: a CLI that turns any repo into an agent-ready workspace. It generates repo-aware agents, skills, workflows, local indexes, and checkpoints. Less giant prompt, more structured agent orchestration. Check it here: github.com/methodnumber13/ru…
1
2
162
afro retweeted
Don't waste 2 years learning to build LLMs like Claude & ChatGPT. Stanford just dropped a 2 hour course on how to build LLMs from scratch. • 00:00 - LLM tokenization • 25:44 - how LLMs decode user prompts • 35:40 - training pipeline of LLMs • 1:16:47 - LLM architecture from scratch Anthropic pays $750,000/year to engineers who understand this exact knowledge of LLMs. Bookmark this & give 2 hours today, no matter what. Then read the article below.
119
1,540
30
7,451
872,559
afro retweeted
这28分钟视频直接把我点醒了。 Anthropic工程师说得很清楚: “你不需要更好的提示词。你需要的是图工程:持久的记忆,这样你的代理永远不会忘记任何事情。” 他完整演示了Anthropic内部到底怎么搭agent、怎么组织工作流。 不是那种堆prompt的玩法,而是真正把记忆做成持久图谱,任务之间能互相引用、互相纠错。 比市面上那些收费agent课强太多了。 强烈建议存下来反复看。
30
258
9
1,017
96,257
afro retweeted
GPT-5.6 Sol, along with Terra and Luna, will launch publicly this Thursday. We’re expanding preview access globally now.
2,414
6,574
4,081
46,780
9,979,409
Another day, another RunWeaver release. v0.3.0 is out: the main focus was making the “init repo → write a task → let the workflow take over” path more reliable across @opencode , Codex(@OpenAI), and @claudeai. github.com/methodnumber13/ru…
1
1
66
What changed: -runweaver-swarm is now the reserved entrypoint across runtimes -OpenCode plugin shadowing detection in `doctor opencode` -terminal workflow state checks before final answers -stronger Codex smoke/live eval support -better task routing and participant selection
1
49
Another day, another RunWeaver update. The new release is out: a more deterministic path from "user asks for a task" to "repo-aware agent swarm starts with the right workflow, context, participants, and checkpoints". github.com/methodnumber13/ru…
1
1
55
What changed in v0.2.0: - `runweaver start` as one stable task-intake command -deterministic workflow + participant selection -task-scoped context retrieval -OpenCode, Codex, Claude adoption checks -MCP context query tool -runtime shortcuts -better checkpoints and resume flow
1
58
RunWeaver v0.1.0 is out. Added repo bootstrap, status/MCP entrypoints, release builds, and richer resume state with pause/reject reason, last result, changed files, stale-index signal, next verification. Thanks @adenweigang for the checkpoint feedback. github.com/methodnumber13/ru…
1
47
RunWeaver update: - repo-aware agent instructions - workflow checkpoints + resume state - stale index detection - OpenCode/Codex/Claude metadata drift checks - optional MCP tools The goal: less giant prompt, more durable repo-native agent orchestration.
1
47
Hey @Aster_DEX! As a community member, I decided to help grow the ecosystem in the way I can. Here’s a TypeScript SDK, now live on npm: 🔗 npmjs.com/package/asterdex-s… Hope it helps both developers and traders⚡ @yzilabs @cz_binance @BNBCHAIN #AsterDEX #BNBChain #DeFi #Web3
2
1
3
223
also tag you guys @Leonard_Aster @Ajay2220111
2
143
Finally done
A bit lower than expected, but still looks good. $Polyagent @MeteoraAG #DLMM
265
A bit lower than expected, but still looks good. $Polyagent @MeteoraAG #DLMM
$Polyagent setup 👇 Farming it on @MeteoraAG DLMM 🔹 CA: FgHpSsku7asGyTThooDmkKwmzy24ucUeyBrW9kLqpump 🔹 Range: 0.00000428437-0.0000084283 Not my play — just sharing the vision. Not FA #Solana #DeFi #Alpha
1
1
442
$qstay also good entry for farming Ca: 96hj47MzSpZgVcT1v4nwpbQNEU59f6ys48PArGivpump @MeteoraAG #DLMM #farming
187
$Polyagent setup 👇 Farming it on @MeteoraAG DLMM 🔹 CA: FgHpSsku7asGyTThooDmkKwmzy24ucUeyBrW9kLqpump 🔹 Range: 0.00000428437-0.0000084283 Not my play — just sharing the vision. Not FA #Solana #DeFi #Alpha
1
1
1
537
It showed up really early in my TG bot.
1
88
Seeing volume go down rn. $Polyagent
175