reviewer: starflinger-anthropic@high + CI validation gate - #2
Merged
Conversation
mc v0.2.0's emitPiModels emits provider.api and model.reasoning from these fields; without them the materialized models.json lacks the anthropic-messages wire protocol and the reasoning flag. api=anthropic-messages (LiteLLM gateway), reasoning=true (minimax-m2.7 supports thinking).
minimax-m2.7 -> starflinger-anthropic. thinking medium -> high (high is the top level pi accepts; pi has no "max" level, and effort=max only fires for opus-4-6 model ids, not this LiteLLM route, so high is the correct ceiling).
The events-ingress container clones this repo @main and parses it with jq on every review, so merge-to-main is the deploy. Add scripts/validate.sh (pure bash+jq, mirrors that contract) and a GitHub Actions workflow that runs it on every PR and push to main, so a malformed agent (e.g. an invalid thinking level) can't reach the live reviewer. Update README open items accordingly.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What
model: starflinger-anthropic,thinking: high(also carries theapi/reasoningfields that PR reviewer: harden prompt for reliable issue-finding + complete model config #1 merged without).highis the ceiling pi accepts — there is nomaxthinking level, and Anthropiceffort: maxonly fires foropus-4-6model ids, not this LiteLLM route.scripts/validate.sh(pure bash + jq) + a GitHub Actions workflow running on every PR and push tomain.Why CI looks like this
The events-ingress container clones this repo
@mainand parses the package withjqon every review — merge-to-main is the deploy, there's no build artifact. The validator mirrors that exact consumption contract (JSON validity, required fields, realthinkinglevel, prompt files exist, toolsets resolve), so a malformed agent can't reach the live reviewer. Catches e.g.thinking: \"max\"before it ships.Closes the README 'mc validate PR gate' open item (implemented via jq, no blocked Zig/mc binary needed).