Agent Information & Policy – The Secret Buddha Advisor
The Secret Buddha Advisor the “Agent” is an AI-powered coach embedded on our site. It can chat, help you plan goals, and — if you consent — remember key points for next time. This page explains how to use the Agent, what data it handles, and your controls.
1) How to Use the Advisor
- Open the chat bubble bottom-right or use the inline chat on supported pages.
- Type naturally, ask questions, or give instructions such as Help me build a morning routine.
- Prefer the inline chat for extended sessions and longer outputs.
2) Commands
Use these in any chat. Commands are case-insensitive.
/help— show this help/export— download your advisor data/remember <type>: <value>— save a fact types: preference, goal, constraint, trigger, habit/facts— list your saved facts top 10/facts remove <needle>— remove the most recent saved fact matching the needle/plan show— display your current plan/plan add <text>— append a step to your plan/plan set <text>— replace your entire plan/plan clear— erase your entire plan cannot be undone/forget <needle>— remove the most recent matching plan line/erase all confirm— erase summaries, facts, and plan cannot be undone
3) PERSONALISATION & Consent
If you are logged in, you can enable memory so the Agent remembers short summaries, your Key Facts, and your Active Plan for continuity. Memory is off by default and requires your explicit consent. You can withdraw consent at any time in the Agent settings ¤ in the pop-up bubble. Without consent or when logged out, the Agent works statelessly and does not remember past chats.
Your controls
- Consent toggle: turn memory on or off in settings.
- Export: download your summaries, Key Facts, and Active Plan as a JSON file.
- Erase: delete all Agent data stored for your account with one click.
- Edit Plan and Facts: update or remove items in settings or via commands.
4) What Data the Advisor Handles
| Data | Purpose | Retention | Your Options |
|---|---|---|---|
| Chat messages | Generate replies to your prompts | Ephemeral for response, not stored as a conversation log | — |
| Summaries optional | Short recap for personalisation | Auto-purged after approximately 180 days default | Export or erase, disable memory |
| Key Facts optional | Stable preferences, goals, constraints | Until you erase them | Add, edit, erase, disable memory |
| Active Plan optional | Your steps and habits | Until you erase it | Edit or erase, disable memory |
| Consent state | Respect your preference for memory | Until changed or account deletion | Toggle on or off anytime |
| Rate-limit metadata | Stability and abuse prevention | Short-lived approximately 10s and 60s | No tracking cookie, keyed to hashed IP |
Note: The Agent does not set additional tracking cookies. It relies on normal WordPress session or auth cookies for logged-in users. Rate-limiting uses short-lived transients keyed to a hashed IP which expire automatically.
5) Third-Party Processing and International Transfers
To generate responses, the Agent may send your prompt and relevant context summaries or facts when memory is enabled to our AI provider currently OpenAI acting as our processor. Processing may occur outside the EU. We apply appropriate safeguards such as standard contractual clauses and the EU–U.S. Data Privacy Framework as applicable. We avoid sending directly identifying data such as your account email in the prompt or context. If the AI service is unavailable, the Agent switches to local Lite mode with no external AI call.
6) Safety, Filtering and Limitations
- Basic keyword triage detects sensitive topics such as self-harm, violence, or medical emergencies and offers supportive, non-clinical guidance and relevant resources.
- The Agent provides general self-improvement guidance. It is not medical or mental-health care and not legal or financial advice.
7) Legal Basis GDPR
- Stateless use without memory: Article 6 1 f legitimate interests to provide the chat service you request.
- Memory features summaries, facts, plan: Article 6 1 a your consent. You can withdraw consent at any time with future effect.
8) Retention
- Summaries: auto-purged after approximately 180 days default.
- Key Facts and Active Plan: retained until you erase them.
- Consent state: retained to honor your preference and timestamp.
- Rate-limit metadata: short-lived approximately 10s and 60s keyed to hashed IP and no cookies.
9) Rate Limiting and Stability
The Agent applies per-IP and per-user when logged in rate limits to keep the service stable. If you exceed limits you may see a brief too many messages notice. Limits reset quickly.
10) Lite Mode Fallback
If the external AI provider is unavailable or returns an error, the Agent switches to Lite mode and provides local, more generic tips. Memory features still follow your consent settings, though replies may be less personalised in Lite mode.
11) No Automated Decisions with Legal Effects
We do not use Agent data to make automated decisions that produce legal or similarly significant effects on you.
12) Practical Guidance for Best Results
- Be specific: describe your goal, constraints, and timeframe. Example: Create a 30-day plan to improve sleep with daily steps of under 10 minutes.
- Iterate: ask the Agent to refine outputs. Example: Make week 2 easier and include two rest days.
- Use inline chat for large outputs: longer plans, summaries, or exports work best in the inline chat.
- Keep sensitive data minimal: do not share passwords, health identifiers, or third-party personal data unless necessary. Prefer generic placeholders.
- Verify important content: treat outputs as guidance and confirm critical facts independently.
13) Troubleshooting
- Export did not auto-download: use the Export button in settings. Some browsers block automatic downloads.
- Hit a rate limit: wait 10 to 60 seconds and resend. Avoid many rapid messages in a short window.
- Scrolling a long reply: the bubble supports vertical scrolling for long messages. For many long replies use the inline chat.
- Unexpected generic answer: rephrase your prompt with more context and examples. If it persists, refresh the page and continue in the inline chat.
14) Disclaimers
- Use your own judgement when following the Agent suggestions.
- The Agent does not replace professional advice medical, mental-health, legal, or financial.
15) Changes and Updates
The Agent may evolve over time. We publish updates here and reflect them in our Privacy Policy. Continued use after updates implies acceptance of changes.
16) Questions or Requests
To access, export, or erase your Agent data or if you have questions contact the person named in our Privacy Policy. You retain all GDPR rights described there.
