← hebel. Today's paper

hebel. today · section

Agentic Engineering.

What 20,000 newsletters said about agentic engineering this week23 stories · 113 sources

Get Agentic Engineering in your inbox every morning

One short email when agentic engineering has stories, every source linked. Stop it with one click.

Sunday 15 November

Coding Agents · 5 sources

Teams move bottleneck from code generation to verification

Software engineering teams are shifting their primary focus from code generation to rigorous verification as agent-driven pull requests now account for over 40% of all submissions. With agents producing code faster than humans can…

AI engineering · 4 sources

Coding agents trade re-planning for reusable workflows

Coding agents historically consume excessive tokens by re-planning every step of a task during each execution. Developers now use Swamp to convert repeated skills into persistent workflows where code manages the execution loop, invoking…

Saturday 14 November

AI Agent Management · 6 sources

Autonomous agent reliability hinges on orchestration versus human oversight

Production teams increasingly favor specialized agent networks over general-purpose systems to ensure operational reliability. Developers at teams using the 2027 standard architecture decompose goals into subtasks, manage dependencies, and enforce fallback strategies through predefined…

AI engineering · 6 sources

Building reliable agent harnesses

Harnessing agent reliability. Engineers now define an agent as a model plus a harness, where the harness comprises a context manager, tool registry, guardrails, a loop, and a verifier. Teams at Fin doubled their…

Agent Evals & Reliability · 5 sources

Production agents stay safe with rigid action catalogs

Horizon3 limited its AI agent to a finite, pre-approved menu of deterministic tools to ensure it stayed within bounds during 420 deployments. This architecture allowed a smaller Claude Haiku 4.5 model to perform as…

Context Engineering · 5 sources

Engineers pivot from prompt engineering to structured component manifests

Design and product teams are shifting from prompt engineering to creating machine-readable manifests that define component properties, variants, and defaults explicitly. By generating these inventories directly from code, teams ensure agents pick from a…

Coding Agents · 4 sources

LinkedIn automates 40,000 weekly pull requests

LinkedIn processes 40,000 pull requests weekly using an AI-native review system that performs an initial pass before human engineers engage. This system addresses the bottleneck created by coding agents, which reduce writing time from…

Friday 13 November

AI engineering · 9 sources

Build production agents or buy off-the-shelf tools

Build for custom workflows. Forward deployed engineers at AI labs integrate bespoke sorters directly into customer systems to ensure production reliability. These engineers treat model swaps as component changes, maintaining agent identity through externalized…

Coding Agents · 5 sources

Teams replace code review with automated intent verification

Verify intent, not lines. Teams no longer review the diff, which shows what changed but never why. At Aviator, engineers now capture intent and acceptance criteria from coding sessions, allowing the same nine-person team…

Context Engineering · 4 sources

Anthropic and agents use 'dreaming' to curate memory

Anthropic models consolidate agent memory through a background process called dreaming, which reviews work logs against existing notes to correct mistakes and prune stale information between sessions. This architecture moves beyond simple vector retrieval…

Thursday 12 November

AI engineering · 5 sources

Deterministic verification outpaces model confidence

Verification over confidence. OpenAI recently processed 2.8 million lines of the Linux kernel to flag 2,308 potential bugs, yet the team discarded most of these after a sandboxed machine failed to reproduce them. Only…

AI Agent Management · 1 source

Autonomous agent coordination creates new management friction

Ethan Mollick argues that management structures for AI agents are becoming obsolete because agents can now coordinate other agents to solve complex problems. He points to a swarm of OpenAI agents that produced a…

Wednesday 11 November

AI Agent Management · 6 sources

Shift from prompting to delegating with structured contracts

Effective agent oversight requires moving away from long, repetitive prompts toward formal delegation contracts that define the assignment's objective, necessary evidence, and required human intervention. While many developers focus on the initial launch, research…

AI engineering · 4 sources

Custom agent harnesses versus framework-based orchestration

The case for custom harnesses. Alexandr Wang at Meta maintained a meticulous spreadsheet of over 100 specific behaviors to validate the Muse agent, blocking any release that failed even 20 of these rows. RevenueCat…

Tuesday 10 November

AI engineering · 5 sources

Decouple decision logic from LLM reasoning

Engineers now offload bounded judgments to specialized decision models to reduce inference costs and improve reliability, effectively splitting agent workflows into a two-tier architecture. Instead of asking an LLM to perform classification or routing,…

AI Agent Management · 4 sources

Design checkpoints replace retrospective review for autonomous agents

Practitioners at Agents4Science found that autonomous agents often fill underspecified briefs with generic, plausible defaults that lack domain-specific rigor. To mitigate this, teams now implement design checkpoints where a human expert approves the agent's…

Monday 9 November

AI Agent Management · 6 sources

Engineering operational guardrails for autonomous agents

Establishing explicit boundaries. Docusign engineering mandates that autonomous agents cannot merge pull requests or execute irreversible actions without human approval. To prevent agents from circumventing operational limits, operators must define an 'Ask-Before-Act' list of…

Your own paper, every morning, for your goals.

Forward the newsletters you already get (a Gmail filter does it) and tell Hebel your goals. Every morning it reads them with 20,000 others and sends one short email, with one small thing to do today. Fewer emails, not more. Free during beta.

Start my paper →