[AI] Edge AI consultation — privacy-gated escalation to cloud LLM
Nobody has claimed this yet.
Assessment
- Difficulty
- 5/5
- Estimated time
- Over a week
- Newbie friendliness
- 25/100
Research direction
Start with CLOUDFEST_HACKATHON.md Goal 3, then inspect the provider abstraction from #21 and the related consultation pattern in #23. Confirm how the WP 7.0 AI Client API could fit. Done means an opted-in, clearly indicated cloud consultation path that strips site-specific data and integrates the cloud response into local reasoning.
Written by the indexing model from the issue text.
Description
Goal
The local LLM acts as a privacy gate — only non-sensitive queries (no PII, no site-specific data) get escalated to a cloud LLM for deeper reasoning.
Use Case
User asks a complex question like "what's the best caching strategy for WooCommerce" or "explain the WordPress hook system". The local model recognizes it needs more knowledge, strips any site-specific context, and escalates to a cloud model via API. The answer comes back and the local model integrates it into its response.
Requirements
- New ability:
consult-cloud-ai - The local LLM decides what to escalate (not the user)
- Site-specific data (URLs, usernames, plugin lists, error messages) must be stripped before escalation
- The cloud response is integrated back into the local reasoning loop
- Clear UI indication when cloud consultation is happening
- User must opt-in to cloud consultation in settings
Technical Considerations
- Builds on the provider abstraction from #21
- Could use WP 7.0 AI Client API when available
- Privacy classification: how does the local model decide what's safe to send?
- Consider a simple heuristic first (no proper nouns, no URLs, no file paths) before LLM-based classification
Context
- See CLOUDFEST_HACKATHON.md Goal 3 — Edge AI Consultation
- Related: #21 (External AI provider architecture)
- Related: #23 (LAN AI — similar consultation pattern but local network)
Skills needed
AI/ML, PHP, React/JS
- Dominant language
- JavaScript
- Stars
- 28
- Forks
- 6
- Avg merge
- 4d 3h
- Merged PRs (30d)
- 2
Getting set up
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from pluginslab/wp-agentic-admin
-
bug
Difficulty 2/5 1-3 hours Newbie friendliness 72/100
pluginslab/wp-agentic-admin#226 ·
-
Difficulty 2/5 1-3 hours Newbie friendliness 65/100
pluginslab/wp-agentic-admin#215 ·
-
WP.org submission: close out review T4 (17 Jun) and stop the review loopMay be free again @ivdimova claimed this 39 days ago, and no pull request is open. Openphp question wporg-submission
pluginslab/wp-agentic-admin#228 · 5 comments · 2 assignees ·
-
enhancement react-js ux
Difficulty 5/5 Over a week Newbie friendliness 35/100
pluginslab/wp-agentic-admin#224 ·
-
enhancement ux
Difficulty 4/5 3-5 days Newbie friendliness 48/100
pluginslab/wp-agentic-admin#223 ·
All issues in pluginslab/wp-agentic-admin
Similar issues
-
curriculum documentation quality
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
githubnext/gh-aw-workshop#3897 ·
Maintainers usually reply within 2 days
-
agent/quality hive/hosted-available-lke648397-260827-5n31 quality testing
Difficulty 2/5 1-3 hours Newbie friendliness 91/100
Maintainers usually reply within 1 day
-
check:failed feeds:add
Difficulty 2/5 1-3 hours Newbie friendliness 70/100
iptv-org/database#36179 · 1 comment ·
Maintainers usually reply within 1 day
-
bug
Difficulty 2/5 1-3 hours Newbie friendliness 88/100
Maintainers usually reply within 1 day
-
Difficulty 2/5 1-3 hours Newbie friendliness 84/100
Maintainers usually reply within 1 day