StartupUnited States
HumanLayer
Human approval and context engineering for agents.
HumanLayer builds tooling for human-in-the-loop agent workflows and authored '12-Factor Agents' — the widely-cited methodology for building production harnesses. Now focused on agentic code review tooling.
Pros & cons
Each point cites its sources and says what kind of evidence backs it. A second reviewer tried to refute every claim before it was kept.
Pros
- MetadataApache-2.0 repo with about 12.7k stars that runs Claude Code, Codex, OpenCode and Pi agents and lets teams steer them from one place.[1]
- DocsBuilds from source send no telemetry; Sentry and PostHog stay off unless configured.[1]
- DocsThe team wrote the widely shared 12-Factor Agents guide, and its CodeLayer product runs parallel agent sessions using worktrees and remote cloud workers.[1]
Cons
- Independent testIt pivoted from a human-in-the-loop approval SDK to coding-agent orchestration, and the original SDK is no longer the main focus.[1]
- DocsSelf-hosting guides are still in progress, and the Docker Compose file ships with default passwords that must be changed.[1]
- MetadataThe repo publishes no pricing, and third-party listings conflict.[1][2]
- DocsIt supports only a few agents (Claude Code, Codex, OpenCode, Pi).[1]
- Best for
- Teams running several coding agents in parallel who want shared steering and visibility.
- Watch out for
- The product is changing quickly after the pivot, so check current docs and pricing before standardizing on it.
Assessed Oct 29, 2026 · adversarially reviewedSuggest a correction
People
Founder & CEO
Context engineering & memory
Their open-source work
TypeScript · ≈27k stars · Other
humanlayer/12-factor-agents
The methodology doc for building production LLM agents.
Maintained by HumanLayer
Activity
- Last push
- · 1y ago
- Latest release
- None published on GitHub
- Last 90 days
- 0 releases
Checked against GitHub on Nov 21, 2026. Release notes are shortened to plain text.
On X
Posts by the company and its team, shown as plain text; “Load on X” loads X's script for that post, which may set cookies.
dex · Nov 20, 2026
dex · Nov 20, 2026
dex · Nov 19, 2026
dex · Nov 17, 2026
Videos

AI Engineer · Dec 2, 2025
No Vibes Allowed: Solving Hard Problems in Complex Codebases – Dex Horthy, HumanLayer
Context engineering & memory · Harness engineering · Talks & interviews
In research (1)
- I-HITL: Inalienable Human-in-the-LoopVerdict: Important
Oct 26, 2026 · Commercial human-approval layer for agents; I-HITL defines the property such approvals must have to be unforgeable by the agent.
Neighbours in the directory
San Francisco, CA · Est. 2023
The integration layer: hundreds of managed tools your agents can call.
Composio handles auth, schemas, observability and reliability for the external SaaS tools an agent uses — 'plumbing for agentic actions' across Slack, GitHub, Salesforce and 250+ other surfaces.
Est. 2024
A registry and toolchain for MCP servers.
Smithery indexes and helps deploy MCP (Model Context Protocol) servers — a distribution layer for the tool ecosystem agents connect to.
San Francisco, CA · Est. 2021
Search API purpose-built for AI agents.
Exa (formerly Metaphor) provides neural search and web-crawl APIs tuned for agent consumption — live, cited knowledge for agents instead of stale training data.
Connections
Relations from the Harness Hub knowledge graph, as charted in the State of Harness reports.
Maintains · 1
- 12-Factor AgentsProject