inferred from GitHub

Agention

OpenClaw power-pack for serious agent execution

Core trick

Packages agent skill sets as installable OpenClaw bundles with a validation suite and upstream sync checks.

Evidence

README matched ""Cerebral Valley" "Opus 4.6"": upstream-sync/work/ecc/docs/zh-CN/README.md

README excerpt

# Agention ⚡ ⚡
[<img src="assets/logo.svg" width="48" height="48">](https://github.com/oabdelmaksoud/Agention)
 for OpenClaw

> **Portable execution intelligence for OpenClaw**  
> Agention brings ECC-style operational patterns, skill orchestration, compatibility tooling, and upstream sync discipline into a practical, installable OpenClaw package.

[![Compatibility](https://img.shields.io/badge/OpenClaw-Compatible-2ea44f)](#compatibility--scope)
[![Skills](https://img.shields.io/badge/Skill%20Pack-ecc--*-0366d6)](#skill-packs-core-vs-full)
[![Validation](https://img.shields.io/badge/Validation-Built--in-orange)](#validation-suite)

---

## Why Agention

Agention is a curated OpenClaw-native distribution designed for teams that want **repeatable agent behavior** with **real operational guardrails**.

### Key benefits

- **OpenClaw-native execution** — no fragile harness emulation; behavior is adapted to OpenClaw surfaces.
- **Large skill inventory** — command/domain/bridge/agency-oriented `ecc-*` skill set in `skills/`.
- **Install + operate + verify lifecycle** — wizard installer, runtime integration, compatibility checks, quality/security/eval scripts.
- **Upstream traceability baked in** — source pins, notices, and sync metadata under `upstream*`.
- **Delegation policy included** — explicit subagent and reporting rules in `SUBAGENT_POLICY.md` and `OPERATIONAL_ENFORCEMENT.md`.

---

## Architecture (ASCII)

```text
                          ┌──────────────────────────────┐

Core trick

Evidence

Tags

README excerpt