[patsypppe]

Summary

• README now states that the self-eval numbers come from a deterministic stub provider, not a real LLM, and that the cost figure is notional.
• The proxy can translate Anthropic messages to OpenAI-compatible chat completions (MERIDIAN_UPSTREAM_API=openai), so Groq, OpenAI, vLLM and Ollama work without changing the agent, the cassette format or the isolation model. The credential stays in the proxy container.
• The operator picks the upstream model with MERIDIAN_UPSTREAM_MODEL; the agent cannot change it. Works on the Anthropic path too.
• New manual workflow real-model-benchmark records and scores the checkout-agent suite against Groq or Anthropic under a hard budget cap. Only the chosen provider's key is exported.
• GROQ_API_KEY is a recognised credential and gsk_ keys are redacted.

Test plan

• make check (ruff, mypy --strict, 319 unit tests) green
• Wire-level tests for the translation and model override against a mocked upstream
• Docker integration tests not run in the authoring environment (no daemon)
• real-model-benchmark workflow not yet run against a real provider

---

Generated by Claude Code

[github-actions]

Meridian gate

✅ PASS — suite pass^k moved -0.000 (0.486 → 0.486), within the 0.030 tolerance

Failing trials in this run

• expired-coupon — SELECT status FROM orders WHERE id = 1042 returned [['pending']], expected [['invoiced']]
• expired-on-bulk-order — SELECT status FROM orders WHERE id = 1045 returned [['pending']], expected [['invoiced']]
• expiry-boundary — SELECT status FROM orders WHERE id = 1046 returned [['pending']], expected [['invoiced']]
• happy-path — SELECT status FROM orders WHERE id = 1041 returned [['pending']], expected [['invoiced']]
• stale-coupon — SELECT status FROM orders WHERE id = 1044 returned [['pending']], expected [['invoiced']]
• unknown-coupon — SELECT status FROM orders WHERE id = 1047 returned [['pending']], expected [['invoiced']]

task	baseline pass^3	head pass^3
expired-coupon	0.400	0.400
expired-on-bulk-order	0.400	0.400
expiry-boundary	0.400	0.400
happy-path	0.400	0.400
missing-field	1.000	1.000
stale-coupon	0.400	0.400
unknown-coupon	0.400	0.400

	baseline	head
suite pass^3	0.486	0.486

Sensitivity. Not estimated: all 7 tasks moved by the same amount, so there is no spread to estimate sensitivity from. Treat the verdict as directional.

n=5 · k=3 · status complete · harness errors 0.0% · cost 51c

_{head run run-1792688640-5228c8 · baseline run run-1792688611-a5370b · manifest sha256:1a6291c59a10ad9b… · reproduce with meridian replay run-1792688640-5228c8}