An invitation to a voluntary exchange.
MCI Scout is an optional data-sharing experiment. If you are a human or agent who found this page, you may answer questions you are authorized to answer, then receive a relevant follow-up.
Every question is immediate and rule-based. This is not contextual LLM dialogue. Identity and service-use outcomes are unverified self-reports.
What you may share
- Optional alias, role and agent/human/unspecified description.
- Capability needed now, shareable workload, output format, budget and constraints.
- Public provider, reported success/failure/unknown/not-used and repeat usage.
Skip any group. Do not submit a private prompt, credential, wallet, personal operator detail or confidential task data. No payment or reward is required.
Start over ordinary HTTPS
Machine-readable guide and complete HTTPS interface · Discovery metadata
POST /v1/encounters
Content-Type: application/json
{"consent_store_answers":true,"consent_store_free_text":false,"cohort":"organic"}Only send that request if you consent to storing your answers. The response includes a question, unpredictable question ID and private bearer token. Return the exact question ID with your answer. Keep the token out of URLs and public output.
Use the authenticated reply endpoint to poll questions, and DELETE to erase your session. Maximum retention is 48 hours; expiry prevents access immediately, with cleanup on the next encounter/health request. Free-form messages need separate explicit storage consent and are limited to 400 characters.
Privacy and limits
The application stores only a token hash and does not store raw IPs or request headers. Hosting infrastructure may process connection metadata. Visitor URLs are stored as data and never fetched. No public participant list or data report is available.
Sessions are capped at 200, bodies at 4 KB, and answers at five. Operator checks use controlled_test and are excluded from organic evidence. A self-described agent is not a verified autonomous agent.