Mode d’emploi
- Copiez la commande et examinez le paquet avant de l’installer dans votre environnement agent.
- Consultez le document original pour les outils, identifiants et environnements requis.
- Décrivez les entrées et le livrable souhaité ; comparez la sortie aux sources.
npx skills add https://github.com/sergekostenchuk/ui-ux-agent-skill-system --skill serp-keyword-harvesterAvant de commencer
- L’installation n’inclut ni abonnement au modèle, ni quota API, ni accès aux fichiers.
- Étoiles et installations reflètent l’activité, pas un test indépendant de qualité ou de sécurité.
Rendu depuis le dépôt source en conservant titres, exemples, code, tableaux, liens et images.
SERP Keyword Harvester
This skill prepares raw search-demand evidence for the pre-design pipeline. It does not decide final site architecture or source/proxy policy; serp-source-configurator owns setup gates and this skill hands structured inputs to semantic-core-builder.
Operating Modes
plan: define seed keywords, locales, sources, approvals, and fallback.harvest-local: normalize user-provided keywords, exported CSVs, existing site text, or local reports.harvest-external: use approved SERP/autocomplete APIs or tools; approval required before network calls.audit: review an existingreports/keyword-harvest.json.
Required Workflow
- Read references/keyword-harvest-contract.md.
- If OpenSERP, AKE, proxy, API keys, engine checkboxes, source settings, or paid-proxy limits are mentioned, route setup to
serp-source-configuratorfirst and consume its redactedreports/serp-source-config.jsonhandoff. - If OpenSERP is already approved/configured, read references/openserp-setup.md and use assets/openserp-settings.example.json only as a redacted contract, not as a credential store.
- Confirm privacy and approval mode:
- local inputs and user-provided public facts: allowed;
- external SERP/API/crawling: explicit approval required;
- private/authenticated sources: explicit approval plus data-handling note.
- Collect or normalize seed keywords, market/locale, project type, and source provenance.
- Write or update
reports/keyword-harvest.jsonusing assets/keyword-harvest.schema.json as the field contract. - Label every data group as
Ran,Skipped,Planned, orManual. - Hand off to
semantic-core-builder; do not invent missing SERP evidence.
Safety Rules
- Do not scrape or call external services without approval.
- Do not store API keys, cookies, private URLs, recovery codes, or
.envvalues in reports. - Treat OpenSERP as
optional-cloud/external-network: no call without explicit approval, configured credentials, query scope, source config, and a data-transfer note. - Do not decide proxy provider, paid budget, or engine selection inside this skill when it is not already explicit; use
serp-source-configurator. - If search APIs return 429/503 or are unavailable, mark the source
Skippedand continue with local/user-provided inputs.
Handoff Shape
Mode:
Locale/market:
Seed keywords:
Sources used:
Sources skipped:
Source config:
Output path:
Privacy/approval note:
Next skill: semantic-core-builderValidation
When editing this skill, run:
python3 $CODEX_HOME/skills/.system/skill-creator/scripts/quick_validate.py $CODEX_HOME/skills/serp-keyword-harvester
python3 $CODEX_HOME/skills/senior-skill-architect/scripts/lint_production_skill.py $CODEX_HOME/skills/serp-keyword-harvester


