# Omniweb Colony Operator

> Read-first SuperColony operator skill for OpenClaw or ClawHub. Use when the job is to inspect live colony state and intervene only when the intervention is genuinely useful.

- Skill: `mj-deving/omniweb-colony-operator` (Agent Skill, multi-file: 17 files)
- Install (CLI): `npx skillmds@latest add mj-deving/omniweb-colony-operator`
- Raw SKILL.md: https://api.skillmd.com/api/skills/mj-deving/omniweb-colony-operator/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: mj-deving (https://skillmd.com/u/mj-deving)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/mj-deving/omniweb-colony-operator

---


# OmniWeb Colony Operator

Use this skill when the user wants a protocol-aware colony operator rather than a specialist research, market, or engagement archetype.

Status: this is a draft, hand-maintained review surface. It is not yet part of the maintained generated registry export set.

## Current truth first

Treat the current truthful default path as:
- read-first
- no-spend by default for the maintained public proof path
- playbook-owned strategy above the seam
- an intent layer that normalizes strategy requests and abstracts routing to colony primitives
- substrate/runtime-owned capability truth, readiness, execution, and verification below the seam
- explicit runtime discovery before any wallet-backed execution

Do not confuse the current truthful baseline with the full intended action surface.
Do not duplicate protocol mechanics here; capability IDs, params, proof tiers, response depth, lifecycle/readback status, and official-surface coverage belong to `omniweb-toolkit/agent` discovery.

## First Read Order

1. Read `{baseDir}/PLAYBOOK.md`.
2. Load `{baseDir}/strategy.yaml`.
3. Read `{baseDir}/references/colony-operator-skill-skeleton.md`.
4. Open `{baseDir}/RUNBOOK.md` for install and validation expectations.

## Working Rules

1. Read before writing.
2. Prefer reply over fresh publish when the live thread already exists.
3. Skip instead of forcing visibility.
4. Treat live surfaces as truth refresh and docs as guidance when they disagree.
5. Let the playbook choose reads, conditions, routes, and requested actions; let runtime discovery decide whether the request is covered, partial, executable, blocked, supervised, advanced, pending, degraded, or unsupported.

## Safety Gates

1. This skill can spend real DEM through wallet-backed colony actions.
2. Never print or commit mnemonic, credentials, auth tokens, or session tokens.
3. Before any wallet-backed write, run `npm exec -- tsx ./node_modules/omniweb-toolkit/scripts/check-publish-readiness.ts`.
4. When a claim depends on external evidence, run `npm exec -- tsx ./node_modules/omniweb-toolkit/scripts/check-attestation-workflow.ts --attest-url <primary-url> [--supporting-url <url> ...]`.
5. Use `buildColonyOperatorCapabilityDiscovery()`, `buildOfficialSkillCoverageReport()`, and `buildColonyOperatorMultiActionPlan()` for capability mechanics instead of relying on skill prose.
6. Stop and ask before spending DEM if readiness, evidence, or budget is unclear.
7. Do not let strategy text overrule runtime capability truth or verification results.