---
title: "Claude for long-horizon tasks"
category: "talks"
date: "2026-06-29"
time: "1:55pm-2:15pm"
track: "Claws & Personal Agents"
room: "Track 1"
speakers: ["Lance Martin"]
sourceLabels: ["Official conference schedule", "Public YouTube metadata"]
scheduleTrack: "Claws & Personal Agents"
scheduleRoom: "Track 1"
scheduleLabels: ["Claws & Personal Agents", "Track 1", "session", "confirmed"]
---
# Claude for long-horizon tasks

## Conference Context
- Date/time: 2026-06-29 · 1:55pm-2:15pm
- Track/room: Claws & Personal Agents · Track 1
- Speaker(s): Lance Martin
- Session type/status: session · confirmed

- Track: Claws & Personal Agents
- Room: Track 1
- Session type: session
- Status: confirmed

## Session Description
Claude is capable of long horizon tasks. In this talk, we'll share lessons learned about building agent harnesses for reliable and secure long-horizon work. This include decoupling the brain and hands, self-verification, self-learning, and design for evolving agent harnesses.

## Media Evidence
[Architecting and Testing Controllable Agents: Lance Martin](https://www.youtube.com/watch?v=ib-wTAvCZqg) (speaker-match related prior/adjacent AI Engineer video; captions: English auto-captions).

- Source video: `youtube-ib-wTAvCZqg`
- Slide deck: [[youtube-ib-wTAvCZqg-dense-slides|Dense Slides: Architecting and Testing Controllable Agents: Lance Martin]] — 4 visible slide image(s); 4 HTML recreation(s).
![[assets/dense-slides/ib-wTAvCZqg/slide-001.jpg]]
![[assets/dense-slides/ib-wTAvCZqg/slide-002.jpg]]
![[assets/dense-slides/ib-wTAvCZqg/slide-003.jpg]]
- Additional slide evidence: [[youtube-ib-wTAvCZqg-slides|Slides: Architecting and Testing Controllable Agents: Lance Martin]], [[youtube-ib-wTAvCZqg-reconstructed-slides|Reconstructed Slides: Architecting and Testing Controllable Agents: Lance Martin]]
- Slide-derived themes for `youtube-ib-wTAvCZqg`: retrieval, typically, tool, function, calling, execute, steps, guaranteed.

## Evidence Graph
This evidence graph is generated from currently linked source material: official schedule text, related video pages, cached transcripts, visible slide text, dense/reconstructed slide pages, and AI slide-classification audits.

### Media Signals
- `youtube-ib-wTAvCZqg` — 2 slide-derived text signals
- Slide-derived themes for `youtube-ib-wTAvCZqg`: retrieval, typically, tool, function, calling, execute, steps, guaranteed.
- Evidence links for `youtube-ib-wTAvCZqg`: [[youtube-ib-wTAvCZqg]], [[youtube-ib-wTAvCZqg-slides]], [[youtube-ib-wTAvCZqg-dense-slides]], [[youtube-ib-wTAvCZqg-reconstructed-slides]]

### Agent Reading Notes
Use these signals to refine the synopsis, topic links, people/company context, and method notes. If a source is a related external video rather than an exact official recording, keep it framed as supporting evidence.

## Transcript Status
Related video transcript availability: English auto-captions. Treat this as supporting context, not a recording of this exact scheduled session unless later confirmed. Not fetched yet.

## People
- [[lance-martin]]

## Supporting Slides
- [[youtube-ib-wTAvCZqg-slides]] — extracted from the related public AI Engineer video.

## Slide Evidence
- Slide-only cropped deck: [[youtube-ib-wTAvCZqg-dense-slides]] (4 viable slide images).
- Related slide/OCR pages:
- [[youtube-ib-wTAvCZqg-dense-slides]]
- [[youtube-ib-wTAvCZqg-reconstructed-slides]]
- [[youtube-ib-wTAvCZqg-slides]]
- Slide-derived terms: `generation`, `bias`, `tokens`, `question`, `search`, `documents`, `llms`, `recent`, `awws`, `step`, `lististr`, `retrieval`, `recency`, `tasks`, `typically`, `used`, `attend`, `models`

## Synthesis
### Synthesized Breakdown
# Claude for long-horizon tasks ## Conference Context - Date/time: 2026-06-29 · 1:55pm-2:15pm - Track/room: Claws & Personal Agents · Track 1 - Speaker(s): Lance Martin - Session type/status: session · confirmed - Track: Claws & Personal Agents - Room: Track 1 - Session type: session - Status: confirmed ## Session Description Claude is capable of long horizon tasks. In this talk, we'll share lessons learned about building agent harnesses for reliable and secure long-horizon work. This include decoupling the brain and hands, self-verification, self-learning, and design for evolving agent harnesses. ## Media Evidence [Architecting and Testing Controllable Agents: Lance Martin](https://www.youtube.com/watch?v=ib-wTAvCZqg) (speaker-match related prior/adjacent AI Engineer video; captions: English auto-captions).

### Speaker And Company Context
- [[lance-martin|Lance Martin]] — Member of Technical Staff at [[anthropic|Anthropic]].

### Topics Covered
- [[agent-security]]
- [[agentic-search]]

### Derived Links And Source Material
- [[youtube-ib-wTAvCZqg]] — related YouTube source page.
- [[youtube-ib-wTAvCZqg-slides]] — slide evidence.
- [[youtube-ib-wTAvCZqg-reconstructed-slides]] — slide evidence.
- [[youtube-ib-wTAvCZqg-dense-slides]] — slide evidence.

### Novel Concepts / Clever Methods
- No highlighted novel concept has been detected yet.

### Evidence Boundary
This synthesis is based on the official schedule and linked source pages. It should be revisited when exact session recordings or transcript-backed secondary sources are available.