The map
The AI Engineering Skills Map is the spine of this deskAndrew Ng’s four skills — wrap, eval, ground, ship — are how this board is filed. If you are building an AI app, start here, not with a model card.
Latent SpaceAug 25Editorial
Setting type
Agentic Firehose
Pulling labs, repos, and the catch-up path…
Harnesses, graphs, loops, APIs you call this week — tagged with Ng’s application skills. Not a training desk. New here is the decoder. Starting a project this week? Catch up is the four-skill map — not more news.
The map
The AI Engineering Skills Map is the spine of this deskAndrew Ng’s four skills — wrap, eval, ground, ship — are how this board is filed. If you are building an AI app, start here, not with a model card.
Latent SpaceAug 25Editorial
A named harness with planning and a filesystem. Compare it to rolling your own before you adopt it.
LangChain2d agoLive
The craft
Harness engineering is the new self-improvement storyLilian Weng’s map of recursive improvement is not sci-fi: it is loops, evals, and the wrapper around the model. If you only watch model cards, you will miss the job.
Lilian WengJul 4Editorial
Cookbooks for building AI apps: context, harness, graphs, cloud.
Wrap
Wrap Claude Opus 5 this week for long-running agentsCallable API drop: coding, computer-use, and long agentic work at Opus price. Swap it in the harness, not a training recipe.
AnthropicJul 24Editorial
Wrap
LangSmith LLM Gateway: runtime controls for production agentsRuntime controls on the LLM gateway for production agents — wrap/ship, not a training launch.
LangChainJul 30Editorial
Wrap
Introducing the Agents API — managed cloud agents on the Codex harnessCallable managed Agents API for orchestration, long-running sessions, and tool use — swap DIY agent loops for a hosted Codex harness path when wrapping/shipping cloud agents.
OpenAISep 10Editorial
Wrap
Gemini 3.8 Flash: SWE/agentic Flash you callProduct drop via Gemini API / AI Studio for long-horizon SWE and agentic workflows.
GoogleSep 2Editorial
Wrap
GPT-6 Astra on OpenAI API (and AWS)Callable Fable-class coding/agent model this week for wrap/eval A/B.
Simon WillisonSep 3Editorial
Wrap
datasette-mcp 0.2MCP server for Datasette so a coding agent can query your tables as tools — wrap the DB you already ship.
datasette/datasette-mcpSep 1Editorial
Wrap
Claude API adds Claude Fable 5.1 with 1M contextSwap the model string in existing Messages API calls for long-running agentic coding on a 1M window without re-chunking context.
Claude API release notesSep 1Editorial
Wrap
Claude for Commerce AgentsReference tool/action surface for shopping and checkout agents you'd wrap around an existing commerce API.
ClaudeSep 2Editorial
Wrap
Wrap GLM-5.3-Flash this week: 1M-context API, not a lab you train atCall it through Z.ai or Baseten — long-context wrap, MIT weights as a swap. Cookbook for the endpoint, not a training drop.
Z.aiAug 26Editorial
App
Next.js App Router: a streaming AI app with the AI SDKBuild the chat UI and the route handler this week. Cloud lab for shipping an application, not wiring a network.
Vercel AI SDKEditorial
Wrap
Call Bedrock Converse in your app this weekWrap the AWS API — messages, tools, streaming — inside the product. Not a VPC lab.
AWSEditorial
Cloud
Deploy a Python ADK agent to Cloud Run this weekGCP lab: Vertex/ADK as an agent app on Cloud Run. Ship the product, not a TPU blog.
Google CloudEditorial
Harness, loops, graphs, MCP, permissions, evals as acceptance.
Harness
Managed agents: decouple the brain from the handsAnthropic’s engineering note on splitting the planner from the tools is the production version of a harness. Same idea whether you like Claude or not.
Anthropic EngineeringEditorial
Clone-to-ship: RAG, eval harnesses, agent wrappers. Not train-a-model.
Repo
earendil-works/piToolkit with agent loop + TUI + coding CLI — drop-in harness pieces for wrapping models with tools instead of rolling your own loop.
GitHub45m ago116k ★Live
Repo
mksglu/context-modeSandbox tool output / shrink context (~98% claimed) for coding agents — drop into harnesses to cut token cost before the
Watchlist posts for people building AI apps. Originals, not architecture notes.
Kubernetes co-creators Craig McLuckie and Joe Beda aim to bring agent harnesses fully into the cloud.
Latent Space10d agoLive
Meet GPT-6.1 Sol: near-Astra intelligence for coding, computer use, and professional work at one-fifth of Astra’s standard API input and output token prices.
browser-use’s Mac harness is the app on the laptop in front of you: Accessibility, a computer-use loop, tool permissions.
browser-use/macos-harnessEditorial
BYOH: MCP and CLI into the loop you already run. An embedded chat agent will not have your skills, memory, or local tools.
yakko.devSep 1Editorial
One command turns a stdio MCP server into a Streamable HTTP URL with OAuth 2.1. Demo the harness without deploying.
mcptunnelsEditorial
Org-provisioned HTTP/SSE MCP for every user — harness permissions, not filesystem MDM.
Claude CodeEditorial
MCPAdapter + elicitation → LangGraph interrupts for app-side agent graphs.
langchain-ai/langchainEditorial
Builder glossary for harness/loop language FDEs ship with.
GitHub BlogEditorial
Redesign tool surfaces for how models call them — harness craft, not an IDE feature list.
Agent ConnectEditorial
Harness
Cursor joins the AI SDK harness layerSwap Cursor, Claude Code, Codex, and Pi behind one HarnessAgent interface — coding-agent wrap without rewriting the app.
VercelAug 27Editorial
Treat agent/env/skill/memory as IaC: describe resources in-repo, run ant apply, commit claude-lock.json so CI updates the same resources instead of recreating them.
Claude API release notesSep 3Editorial
Faster cold paths for coding-agent / cloud-agent sandboxes — ship agent loops that spawn remote sandboxes without the previous routing tax.
Vercel ChangelogSep 8Editorial
Server-side auto allow/deny/pause for agent + MCP tool calls (evaluation field on events) plus terminal attach via ant beta:sessions connect — ship Managed Agents harnesses with live approval loops.
Claude Platform release notesSep 10Editorial
Same-laptop bake-off across nine coding harnesses — use it to pick harness by local cost×quality before you standardize a fleet.
Hacker News / NotionSep 10Editorial
Gotchas moving huge agent preprompts onto self-hosted Ollama
patrickmccanna.netSep 14Editorial
GitHub20m ago30k ★Live
Repo
ChromeDevTools/chrome-devtools-mcpOfficial-ish MCP surface so coding agents can drive DevTools — ground browser/debug tool use in harnesses without scraping CDP by hand.
GitHub21h ago54k ★Live
Repo
bytedance/deer-flowOpen long-horizon agent harness with sandboxes/memory — clone research+code+create loops instead of writing orchestration from scratch.
GitHub1h ago84k ★Live
Repo
slowave-ai/slowaveLocal adaptive memory harness for coding agents — keep context without a cloud store
GitHub2d ago17 ★Live
Repo
anomalyco/opencodeActively pushed open coding-agent harness — clone-to-ship an alternative terminal agent loop without a closed SaaS lock-in.
GitHub2h ago215k ★Live
Repo
addyosmani/agent-skillsClone a curated skill pack for coding agents — drop production engineering skills into Claude Code/Cursor/Codex-style harnesses instead of inventing prompts from scratch
GitHub4d ago108k ★Live
Repo
upstash/context7Up-to-date docs for LLMs and coding agents — ground harnesses on current library docs instead of stale training cutoff.
GitHub12h ago63k ★Live
Repo
DeusData/codebase-memory-mcpMCP server that indexes repos into a persistent code graph — clone-to-ground coding agents with structured codebase memory.
GitHub1h ago47k ★Live
Repo
deepmemteam/deepmem-claude-pluginMemory plugin so coding agents remember coding & chat context across sessions
GitHubAug 202 ★Live
Repo
anhermon/awesome-agent-observabilityMaps: curated list of agent observability tools, standards and platforms (OTel GenAI, tracing, evals, guardrails, MCP) Clone-to-ship: RAG, eval, or an agent wrapper.
GitHub57m ago52 ★Live
Repo
MemPalace/mempalaceClone-to-ship agent memory layer (benchmarked OSS) — persist cross-session context for RAG/agent loops without a paid memory SaaS.
GitHub19m ago60k ★Live
X
Orchestrators vs graphs: the ceiling of a central routerHarness craft: an orchestrator routes every message; graphs let agents hand off without a single bottleneck reading all five.
Hanako / XSep 14Live
X
Altman: stop writing prompts — build loops and graphsAgent craft spine: Prompts → Agents → Loops → Graphs — the difference between using AI and having AI work.
virgilxbt / XSep 14Live
X
Open Code Review: deterministic pipeline + agent only for reasoningCoding-agent craft: Git diffs → review changed files → codebase context search — agent where judgment helps, not everywhere.
Tech with Mak / XSep 14Live
X
ElevenLabs MCP: voice/music/image/video from the assistant you already useMCP wrap craft: generate speech, dubs, SFX, images, video from the coding/chat agent — tools in the harness, not a new app.
Charly Wargnier / XSep 14Live
X
Book drop: Agentic Architectural Patterns (multi-agent + RAG + LLMOps)Catch-up systems: design patterns for multi-agent systems, RAG, and enterprise-scale AI app architecture — not model training.
Kirk Borne / XSep 14Live
X
Anthropic: teams of agents with loops, graphs, skills & hooksHarness course: CLAUDE.md + Plan mode → skills/hooks → subagents → self-improving loops — app-builder spine.
Callan / XSep 14Live
X
Claude Mods land: extend the coding agent with community modsWrap/harness: Mods let you extend Claude’s agent surface — technical details + demos in the community update.
Boris Cherny / XSep 14Live
X
"End of Prompt Engineering": wishes aren't harnessesAgent craft: telling agents what to do is finished — loops/graphs beat repeating the same wish 300 times.
virgilxbt / XSep 14Live
X
Anthropic: production Claude Code harness course (CLAUDE.md + Plan Mode)Harness craft: 1-hour on how Claude Code harness works — long-horizon agents via CLAUDE.md & Plan Mode.
Lunar / XSep 14Live
X
Google Graph & Loop course (fresh drop): 1 prompt → 100 agentsGraph engineering catch-up: first agent graph, parallel agents, loop routing, self-improving graphs — shape over prompts.
Hanako / XSep 14Live
X
Coding agent ships CLI + writes/runs its own unit testsEval/ship craft: one-line Codex task → script, mock test, then self-authored unit tests that pass — not a demo script.
Shraddha Bharuka / XSep 14Live
X
Encode your job into Claude Skills — automate the daily workflowHarness craft: Anthropic paper on packaging daily workflows as custom Skills so the agent does the heavy lifting.
Annatar.md / XSep 14Live
X
Interview: coding agent plans a PR — then a human overlaps commitsFDE agentic engineering: how does your agent reconcile overlapping changes mid-plan-to-PR? Resume keywords aren't enough.
avrl / XSep 14Live
X
Workflows vs graphs: fixed boxes hit a ceilingGraph craft: a workflow decides every step before it runs — six months later the same three boxes fire. Graphs adapt.
Hanako / XSep 13Live
X
20-min crash course: build→test→deploy→evaluate→improve via coding agentCatch-up spine: full agent lifecycle in plain English through Codex — ship loop, not model internals.
Charly Wargnier / XSep 13Live
Harness
How to design an agent harness: work inside the system of recordAgents share data and policy with humans in one continuously updated view, with an audit trail. Harness design, not a valuation hook.
Yarchi / XAug 27Editorial
Graph
Ng graph engineering: loops into working graphs (2h)Agentic knowledge graphs as app craft: first agent, how loops work, freeze them into graphs. Not a training lecture.
Anatoli Kopadze / XAug 27Editorial
Course
Hugging Face agents course: ship an agent, freesmolagents, LlamaIndex, LangGraph, then a project with evals. Catch up already files the lab; this is the pointer.
Ren / XAug 27Editorial
Graph
Ng’s agentic knowledge graphs course (1h)Persistent context for agents you ship: build the graph, then the multi-agent system. Not a grind thread.
Swati Gupta / XAug 26Editorial
Harness
Claude Code ELI5 skill: MCP and harness as a pageA skill that draws the wrapper: MCP runtime and harness, not a model card. Install it on a real ticket.
GoFly / XAug 25Editorial
Graph
Memory vs graphs: a pile vs a routeMemory is continuity. The graph is who may retrieve and write back. Routing, not a bigger embedding store.
Lunar / XAug 25Editorial
Lovable’s CTO on shaping the build: ship tools, not another consumer app. MCP is the interface.
Latent SpaceAug 25Editorial
Loop
Loops vs graphs: agent owns the path, or you draw the nodesLoop while the work is messy. Freeze it into a graph once the pipeline repeats. Same company brain either way.
Annatar.md / XAug 24Editorial
Loop
Karpathy: agents → loops → graphs as the app craftThe lecture is wrap/eval/ship: loops and graphs you run, not how a lab pretrains a model.
virgilxbt / XAug 23Editorial
Claude
Full Claude Course: practical Claude for shipping an appWrap Claude, add tools, ship a real project. Practical course, not a hustle bait.
Jobescape / XAug 11Editorial
FDE
No-BS FDE guide: wrap and ship in the customer’s stackForward-deployed craft: you are in the account building the AI app. Not a salary-in-30-days thread.
Rahul / XAug 1Editorial