인터페이스는 현지화되어 있으며, Skill 소개와 설명 문서는 선택한 언어로 AI가 자동 번역합니다. 코드·명령어·경로·전문 용어는 원문을 유지하며 원작자 문서가 우선합니다.
Skills 둘러보기
생산성 및 협업 · SKILL.md

harvesting-and-admin

harvesting-and-admin

Operate GeoNode harvesting, indexing, asynchronous workers, management commands, thesauri, migrations, and backup/restore workflows with explicit service gates and recovery safeguards.

Independent GitHub snapshot · Automated classification · Original documentation governs

Skill 문서

GitHub

name: harvesting-and-admin description: "Operate GeoNode harvesting, indexing, asynchronous workers, management commands, thesauri, migrations, and backup/restore workflows with explicit service gates and recovery safeguards." disable-model-invocation: true metadata: disco-role: operating license: NOASSERTION

Harvesting and administration

Use this sub-skill when a task involves remote catalogue harvesting, search index maintenance, Celery or async execution, Django management commands, admin operations, thesauri, migrations, or backup/restore planning. It is an operational router, not proof that a GeoNode deployment or its external services are available.

Safety boundary

  • Treat PostgreSQL/PostGIS, GeoServer, Redis/broker, Celery workers/beat, remote OGC or ArcGIS endpoints, credentials, writable media/data directories, and browser/admin services as explicit prerequisites. None is verified by a Python import or by this skill's help wrapper.
  • Prefer inspection and --help. Never use --force, --skip-read-only, restore, deletion, overwrite, or network harvesting as a default.
  • Before any mutating action, identify the target instance, take a tested backup, narrow the selection, record the command and expected row/resource count, and arrange rollback or recovery. Do not run production changes just to validate a command.
  • Keep secrets out of shell history and logs. Do not pass passwords on command lines when an environment-specific secret mechanism is available.

Route the request

  1. Harvesting or a stalled asynchronous operation: read harvesting-and-async.md, then troubleshooting.md. Separate remote URL, database, broker, worker queue, and GeoServer gates before retrying.
  2. Command selection, migration, reindex, sync, or thesaurus: read management-commands.md. Discover syntax with the bundled help wrapper; classify help-only, local-mutating, destructive, and service-backed operations before execution.
  3. Backup or restore: read backup-restore-and-safety.md and the recovery section of troubleshooting.md. Validate archive/config/space and the recovery file before considering a restore. A restore overwrites the target by default.
  4. Unknown failure, stuck state, or partial result: use the decision tree in troubleshooting.md, preserve logs and session/job identifiers, and stop at the first unverified service gate.

Minimal operating sequence

  1. Record the GeoNode release/snapshot, settings module, target environment, operator permissions, and whether this is a disposable copy.
  2. Run python scripts/manage_help.py --help to inspect the wrapper, then use --project-root with --command-help COMMAND or --list-commands. The wrapper is help/list only; it does not execute an arbitrary command.
  3. For a management command, run its help first, inspect the source/documented preconditions, select the narrowest filter and dry-run option, and capture stdout/stderr. The wrapper cannot make a mutating command safe.
  4. Verify expected observations: status transitions to ready, a session/job reaches a terminal state, selected index rows are rebuilt, archive and MD5 files exist, or a migration/check reports success. Confirm the actual DB, catalogue, and file effects separately.
  5. If a step fails, do not blindly rerun. Record the command, options, session or job id, queue/worker, traceback, and last successful stage; use the recovery guidance before retrying.

Acceptance gates

A result is acceptable only when the command's side effects and prerequisites are classified. Help output alone proves parser availability, not database, GeoServer, broker, remote endpoint, permission, or data correctness. For harvesting, require remote availability plus completed refresh/harvest session signals. For indexing/sync, require the correct database and, where used, GeoServer/catalogue services. For backup, verify archive integrity and a separate recovery plan. For restore, use an isolated target and verify fixtures, media/assets, GeoServer state, and the recorded restore history.

Bundled helper

scripts/manage_help.py accepts an explicit project root or an installed manage.py/entrypoint, lists commands, or prints one command's help. It rejects command arguments other than its own safe list/help modes and never runs a mutating management command. See the command matrix and examples in management-commands.md.

Scope and links

This sub-skill owns harvesting/indexing, async/Celery administration, management-command safety, thesauri, migrations, backup/restore, and operational recovery. Route installation and service provisioning to the setup sub-skill; route resource/API payload mechanics to the resource/API sub-skill; route GeoServer authentication and publication internals to the GeoServer and security sub-skill. Those boundaries do not remove the service gates stated here.

저장한 Skills