LangChain lays out the 5 components of a "company harness" for internal agents
hwchase17 · x · 2026-10-28
LangChain founder hwchase17 published an article on building a "company harness," drawing on internal agents his team has deployed across marketing, sales, and engineering. Sydney Runkle's summary: coding agents were the first agents with real impact, but company agents need different primitives —
- UX: agents live where you work (e.g. Slack)
- Identity: agents can act on behalf of the user or with a fixed identity
- Memory: remembering preferences and feedback across sessions
- Sandboxes: safe ways for agents to write code
- Schedules: runs can be scheduled by users and by agents
The piece covers how these components shaped LangChain's internal agent deployments.
More from coding & agent
- AI employee Viktor hits $50M ARR about seven months after launch — SucceededMind · 2026-10-28
- Row-Bot: a local 27B model runs a shop's morning ops entirely on one PC via MCP — Acceptable-Object390 · 2026-10-28
- Offboarding AI agents: revoke sessions first, deleting the container isn't enough — ComprehensiveShake76 · 2026-10-28
- The compounding error problem: 95% per-step accuracy means 36% over 20 steps — KelvenSevero0 · 2026-10-28
- Dev shares AI game project's review list: laser bugs, podracer handling fixes queued — jasonkneen · 2026-10-28
- Dev builds online multiplayer castle-storming game with GPT 6.1 Terra, playable on mobile — chongdashu · 2026-10-28