Skip to content

feat: add TinyHumans managed model adapter - #18

Merged
senamakel merged 1 commit into
epic/8092-inference-outfrom
epic-8092-u8-managed-model
Oct 18, 2026
Merged

senamakel merged 1 commit into
epic/8092-inference-outfrom
epic-8092-u8-managed-model

Conversation

@senamakel

Copy link
Copy Markdown
Owner

Move the TinyHumans managed OpenAI-compatible model into TinyInference. The adapter resolves its bearer for every request while retaining a pooled HTTP client, shapes reasoning and explicit thread fields, projects the managed usage envelope on unary and completed streaming responses, reports authentication failures with call-operation context, and exposes authenticated catalog discovery. The host continues to supply credentials, endpoint policy, HTTP client, and attribution headers.

Coverage includes loopback checks for token rotation, resolver failures, request model selection and fallback, auth failures at dispatch and inside HTTP 200 streams, stream usage projection, readiness classification/fail-open behavior, request shaping, capabilities/cache identity, and authenticated catalog listing.

Validation: cargo fmt --all -- --check passed (evidence: /home/enamakel/work/openhuman/.git/implementing-an-epic/evidence/ti8-format-check.log). Local Rust package test/check/clippy gates are pending release of the shared build slot; hosted CI will run on this PR head.

Refs #8
Part of tinyhumansai/openhuman#8092

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant