docs: fix the guidance-validation runbook — MCP_URL is what points a gate run at dev - #351
Merged
Merged
Conversation
…gate run at dev The recipe named geo-agent-template as the dev-targeted app, but its mcp_url now points at the cirrus mirror head, so following it measured prod guidance — not the change under test. The runner also had no MCP override at all (added in open-llm-proxy#100/#101); document MCP_URL as the knob, and make checking the Job log's mcp line part of the read-back. Also replaces the "seed a standing baseline set" plan with a pointer to geo-agent-benchmark, where the regression tier, gold answers, tiers, grading, and trap tags now live, and updates the model-registry pointer.
This was referenced Aug 8, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Found while gating #312 (PR #350): the Validating guidance changes runbook could not be followed as written.
geo-agent-templatebecause itslayers-input.jsonpointsmcp_urlat dev. It doesn't any more — it points atduckdb-mcp.carlboettiger.info(the cirrus mirror head). So the documented recipe silently measured prod guidance, i.e. not the change under test.MCP_URLknob and makes "check the Job log's--- mcp: ---line" part of step 3.boettiger-lab/geo-agent-benchmarkas theregressiontier with gold answers, tiers, mechanical grading, andtrap:tags — so the section now points there instead of describing a plan.geo-agent-template/k8s/configmap.yaml; the registry is nowgeo-agent-benchmark/suite/models.yaml.Also notes that reading
PROXY_KEYout of the cluster Secret onto a local disk is not the supported path — the Job mounts it in-pod, which is the reason the Job path exists.No prompt artifact changes, so no model-suite validation applies to this PR (
AGENTS.mdis not injected into any tool description).