Authority profile · compiled by Lazyweb
Lance Martin
Anthropic engineer and speaker on long-horizon Claude agent harnesses.
Lance Martin works on agent systems at Anthropic and speaks publicly about harnesses for reliable, secure long-horizon work. His talks connect model capability to the execution environment around it.
Original profile or publication
Role comes from the linked Anthropic talk description; portrait comes from the linked public GitHub profile. Checked September 19, 2026.
- Updated
Referenced in 101
- Managed agents 101
Claude for Long-Horizon Tasks
Long-horizon reliability comes from the harness and environment around Claude: decomposition, state, feedback, security and recovery.
About this page
This is a Lazyweb reference profile, not an account, endorsement or claim that Lance Martin contributed to Lazyweb. The profile identifies an author, editor, maintainer, speaker or source publication connected to the linked work; it is not necessarily the article byline. Summaries are Lazyweb's; the linked work remains the source of truth.