# vectorize-io/self-driving-agents

Repository: https://github.com/vectorize-io/self-driving-agents
Canonical: https://ross.abutalabs.com/products/self-driving-agents
Homepage: https://selfdrivingagents.ai/
Language: TypeScript
License Family: other
Last push: 2026-08-26T20:25:36+00:00

## Health v2 (maintenance only)
Score: 78/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 89, longevity 9
- inputs: {"age_days": 127, "days_push": 7, "days_rel": 79, "gap_med": 0, "n_releases_24m": 18}
- flags: young, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2730, forks 806 (observed 2026-08-28T04:07:16.053192+00:00)

## What it is
A collection of 179 ready-to-use, self-learning AI agent definitions organized into 13 departments (design, engineering, marketing, sales, testing, etc.), installable via an npm CLI into harnesses like Claude Code, Claude Chat, OpenClaw, and NemoClaw. Agents retain durable memory through Hindsight, auto-updating their own knowledge pages after each conversation so they improve over time.

## Use cases
- install prebuilt AI agents for marketing, SEO, or engineering into Claude Code
- give my AI assistant persistent memory across sessions
- find a collection of specialized AI agent prompts for design and UX work
- build a custom AI agent that learns from conversations
- set up an AI workforce covering sales, support, and product departments
- create agents that maintain their own knowledge pages and playbooks
- install agents from a local directory or GitHub repo into a chat harness

## When to choose
- you use Claude Code, Claude Chat, OpenClaw, or a compatible harness and want curated, domain-specific agents
- you want agents with persistent, self-maintaining memory powered by Hindsight
- you need a broad roster of role-based agents (design, finance, marketing, testing) without writing prompts from scratch

## When to avoid
- you need a production agent runtime or orchestration backend rather than agent definitions and an installer
- you work outside the supported harnesses (e.g., custom LangChain pipelines or other LLM tooling)
- you require a license file or formal support guarantees — the repo has no explicit license listed in metadata

## Facets
- artifact type: library
- maturity: active
- function: agent-framework, cli, prompt-engineering, chatbot
- domain: large-language-models, developer-tools
- platform: cli, cross-platform
- tags: ai-workforce, agent-templates, agent-memory, hindsight, claude-code, self-learning-agents, npm-package, mit-license, ai-agents, automation, nodejs

## Member repositories
- vectorize-io/self-driving-agents (main) score 78

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:16.053192+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:14:41.450189+00:00, confidence not recorded.
  - readme: https://github.com/vectorize-io/self-driving-agents (fetched 2026-08-28T04:07:16.053192+00:00, sha 251341fc8c37)
  - homepage: https://selfdrivingagents.ai/ (fetched 2026-08-29T09:58:00.439793+00:00, sha f0ee199aa4f9)
  - site_page: https://selfdrivingagents.ai/agents/engineering/ai (fetched 2026-08-29T09:58:00.449345+00:00, sha d60719a4fedc)
  - site_page: https://selfdrivingagents.ai/agents/engineering/backend (fetched 2026-08-29T09:58:00.453627+00:00, sha 451748f32c75)
- Data as of 2026-08-30T08:39:29.467469+00:00.