Allow configuring how long the worker model stays resident on the Ollama
target after a request via FOREMAN_KEEP_ALIVE env var. Accepts Ollama
duration strings ("-1" forever, "0" unload, "15m", "1h", etc). Defaults
to "-1" (pin forever). The embedder warm-up is unaffected and always
uses keep_alive=-1.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds client/ -- a public Go package providing a synchronous facade over
foreman's async POST /jobs API (Level 1 integration per ADR-0011).
Two delivery modes:
- Webhook receiver (preferred): ephemeral HTTP server on random port,
pushes results immediately, verifies HMAC when configured
- Polling fallback: polls GET /jobs/{id} at configurable interval
Also includes Tags() and Embed() helpers, bearer auth support, and
comprehensive integration tests against the real foreman HTTP handlers.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
steve