Skip to content
Back to skills

Gis Admin Content

ASecurity

"Operate ArcGIS GIS authentication, content, item resources, users,

  • 247 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added September 8, 2026
securityapi

Works with

  • cli
  • api

Security analysis

A100/100

Pro scans all 4 files and shows the line behind each finding

Scanned September 8, 2026

npx -y skills add VectorSpaceLab/AREX-Skill --skill gis-admin-content --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Gis Admin Content?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Gis Admin Content
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/vectorspacelab-gis-admin-content/badge)](https://www.skillsdirectory.com/skills/vectorspacelab-gis-admin-content)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: gis-admin-content
description: "Operate ArcGIS GIS authentication, content, item resources, users,
  groups, organization administration, cloning, and safe backup workflows."
disable-model-invocation: true
metadata:
  disco-role: operating
license: Apache 2.0
---

# GIS administration and content router

Use this sub-skill when the task involves the ArcGIS `GIS` object, authentication profiles, content and item lifecycle, item resources, user/group administration, organization settings, federated servers, collaborations, portal cloning or offline content backup.

Do **not** use this sub-skill for feature-layer analysis, Spatially Enabled DataFrames, mapping widgets/location services, raster/imagery analytics, deep learning, StoryMap/Experience/Knowledge-specific edits, or deployment packaging. Route those to the matching sibling sub-skill or the root skill.

## Operating rules

- Treat all authenticated ArcGIS Online/Enterprise calls as remote service calls. They are not safe to execute unless the user supplies the target portal/org, credentials or profile, required privileges, and explicit approval for the operation class.
- Never paste, log, persist, or echo passwords, tokens, certificate passwords, API keys, client secrets, or profile contents. If a password is needed interactively, prompt through secret input or ask the user to configure a profile outside the response.
- Keep `verify_cert=True` by default. Only use `verify_cert=False` for a controlled diagnostic against a known non-production/self-signed endpoint after warning about man-in-the-middle risk.
- Before any create, publish, share, clone, credit, server, collaboration, reassign, unprotect, or delete operation, perform the preflight and rollback checklist in [admin-content-workflows.md](references/admin-content-workflows.md).
- Do not run or copy bundled portal-population, cleanup, setup, teardown, or clone samples as runtime helpers. They are credentialed and destructive; use only the distilled safety patterns in the references.

## Reference map

- [Admin/content workflows](references/admin-content-workflows.md): connection modes, safe content CRUD, resources, users/groups, admin/server/collaboration, cloning, offline backup, and source-script safety boundaries.
- [API reference](references/api-reference.md): verified installed signatures and common object/method patterns for `GIS`, `Item`, content, resources, users, groups, admin, servers, and collaborations.
- [Troubleshooting](references/troubleshooting.md): missing credentials, profile/keyring failures, SSL/cert issues, privilege errors, clone/delete failures, offline backup issues, and recovery after partial mutations.

## First response pattern

1. Identify whether the user is asking for read-only inventory, reversible mutation, or destructive/admin mutation.
2. Ask for the minimum missing portal/profile/credential/privilege details; if missing, provide a dry-run plan only.
3. Use [api-reference.md](references/api-reference.md) for exact constructor and method parameters.
4. Use [troubleshooting.md](references/troubleshooting.md) when authentication or portal calls fail.
5. For destructive operations, require a human-readable target list, backup/rollback plan, and explicit confirmation before proceeding.

Files in this skill

  • SKILL.md3.2 KB
  • references/admin-content-workflows.md12.3 KB
  • references/api-reference.md11.6 KB
  • references/troubleshooting.md7.6 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.