Conversation
LC-velocity classifier improvements. Bundles two compile-worker prompt classifier fixes that surfaced from mmnto-ai/liquid-city authoring friction. Scope: - .changeset/1.15.4-lc-velocity-classifier-improvements.md: patch-level across the fixed group. Two-section release notes covering the test-contract scope classifier (#1626 via #1652) and declared severity override (#1656 via #1658), plus strategy submodule bump context. - docs/active_work.md: preamble rewritten for 1.15.4. Recently Shipped entry covers both PRs with SHAs and per-feature detail, the LC-queue state pre/post-merge, and the bot-review tail on #1658 (four rounds: GCA R1 + CR R1 major parser strictness, CR R1 info schema examples, GCA R2 stripInlineCode order, CR R3 minor rejection-path severity override). Counts unchanged from 1.15.3 since neither PR touched rules or lessons. - .strategy submodule bump: 113179c -> 7892892b. Picks up strategy PR #125 (upstream-feedback items 015 + 016 from LC session-17, filed as totem#1656 + totem#1657) and strategy PR #124 (upstream-feedback item 017, the three-layer language-support-gap addendum documenting the alpha + beta + gamma architectural pattern for the pending Rust-support arc). Downstream impact: - #1626 fix: liquid-city test-contract rules ("Normalize temp paths for cross-platform equality", "Spy on logger contracts in tests") previously shipped with scopes that excluded tests and silently never fired. Retriage via totem compile --upgrade closes this on a follow-up chore PR. - #1656 fix: liquid-city ADR-008 authoring burned ~10 manual severity edits across R2 + R3 rounds because the compile pipeline emitted warning despite lesson prose declaring Severity: error. The mechanical re-edit loop goes away. No code changes in this PR; the two feature PRs already landed on main with their own test suites passing. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Summary of ChangesHello, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed! This pull request performs a patch release (1.15.4) for the compile-worker, focusing on improving the fidelity of compiled rules based on lesson prose. By introducing a new test-contract scope classifier and a mechanism for declared severity overrides, the changes ensure that compiled rules better align with the author's original intent, particularly regarding test file coverage and severity levels. These improvements resolve specific authoring friction points and include necessary documentation updates and strategy submodule synchronization. Highlights
Using Gemini Code AssistThe full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips. Invoking Gemini You can request assistance from Gemini at any point by creating a comment using either
Customization To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a Limitations & Feedback Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for GitHub and other Google products, sign up here. Footnotes
|
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (3)
📜 Recent review details⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (3)
🧰 Additional context used📓 Path-based instructions (4)**/*.{md,sh,yml,yaml}📄 CodeRabbit inference engine (.github/copilot-instructions.md)
Files:
**/*.md📄 CodeRabbit inference engine (.github/copilot-instructions.md)
Files:
**/*.{md,mdx,txt}📄 CodeRabbit inference engine (.github/copilot-instructions.md)
Files:
**/.changeset/**📄 CodeRabbit inference engine (GEMINI.md)
Files:
🧠 Learnings (28)📓 Common learnings📚 Learning: 2026-03-21T03:31:30.520ZApplied to files:
📚 Learning: 2026-03-27T22:48:00.584ZApplied to files:
📚 Learning: 2026-03-30T05:03:53.061ZApplied to files:
📚 Learning: 2026-04-09T05:03:27.048ZApplied to files:
📚 Learning: 2026-04-10T17:31:50.739ZApplied to files:
📚 Learning: 2026-04-18T04:10:57.267ZApplied to files:
📚 Learning: 2026-03-26T23:36:46.435ZApplied to files:
📚 Learning: 2026-04-18T19:32:25.001ZApplied to files:
📚 Learning: 2026-03-25T00:50:00.878ZApplied to files:
📚 Learning: 2026-04-23T00:38:42.085ZApplied to files:
📚 Learning: 2026-04-03T17:10:29.834ZApplied to files:
📚 Learning: 2026-04-23T20:02:57.991ZApplied to files:
📚 Learning: 2026-04-17T02:53:39.654ZApplied to files:
📚 Learning: 2026-03-25T02:28:36.893ZApplied to files:
📚 Learning: 2026-03-24T22:27:12.418ZApplied to files:
📚 Learning: 2026-04-11T23:20:23.016ZApplied to files:
📚 Learning: 2026-03-26T23:56:46.524ZApplied to files:
📚 Learning: 2026-03-27T05:26:52.361ZApplied to files:
📚 Learning: 2026-03-26T01:55:20.142ZApplied to files:
📚 Learning: 2026-03-26T01:55:25.072ZApplied to files:
📚 Learning: 2026-04-12T15:58:52.811ZApplied to files:
📚 Learning: 2026-04-23T19:46:46.138ZApplied to files:
📚 Learning: 2026-03-22T18:56:51.482ZApplied to files:
📚 Learning: 2026-03-27T20:44:18.105ZApplied to files:
📚 Learning: 2026-04-06T17:10:32.088ZApplied to files:
📚 Learning: 2026-04-13T03:12:22.564ZApplied to files:
📚 Learning: 2026-04-12T21:16:14.364ZApplied to files:
🔇 Additional comments (7)
📝 WalkthroughSummary by CodeRabbit
WalkthroughAdds two classifier improvements for version 1.15.4: a Test-Contract Scope Classifier teaching the compiler to detect and emit test-inclusive fileGlobs for hazards in test files, and a Declared Severity feature that parses prose severity declarations and can override compiled severity with deterministic logic and telemetry tracking. Updates .strategy submodule pointer. Changes
Sequence DiagramsequenceDiagram
participant Compiler
participant LLM
participant RuleBuilder as Rule Builder
participant Telemetry
Compiler->>LLM: Request compiled rule (with prompt)
LLM-->>Compiler: Return severity + lesson body
Compiler->>RuleBuilder: buildCompiledRule(lesson, declaredSeverityOverride)
RuleBuilder->>RuleBuilder: parseDeclaredSeverity(body)
Note over RuleBuilder: Extract prose severity<br/>("**Severity:** error|warning")
RuleBuilder->>RuleBuilder: Apply override if severity changed
RuleBuilder-->>Compiler: Return compiled rule + severityOverride marker
Compiler->>Telemetry: onSeverityOverride callback<br/>(if override changed result)
Telemetry->>Telemetry: Append severity-override<br/>record to .totem/temp/telemetry.jsonl
Estimated Code Review Effort🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly Related PRs
Poem
🚥 Pre-merge checks | ✅ 3 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (3 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request prepares the 1.15.4 patch release, introducing two major compile-worker improvements: a Test-contract scope classifier (PR #1652) for better test file globbing and a Declared severity override (PR #1658) to honor prose-based severity declarations. The changes also include a strategy submodule bump and updates to the active work documentation. I have no feedback to provide.
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.
Summary
Patch release bundling two compile-worker prompt classifier improvements that surfaced from
mmnto-ai/liquid-cityauthoring friction. Both close fidelity gaps between lesson prose authors wrote and the compiled rule that shipped.3037591d) — Test-contract scope classifier (closes core: compile-worker scope derivation mis-scopes test-contract lessons away from tests #1626). Compile-worker recognizes test-contract lessons via three signals (testingtag, test-framework calls in examples, lesson-body references) and emits test-inclusivefileGlobsinstead of the default!**/*.test.*exclusion. Anti-lazy guard against "API Contracts" / "Data Contracts" keyword-only false positives.66e26da2) — Declared severity override (closes core: compile LLM emits 'severity: warning' even when lesson prose declares 'Severity: error' #1656). NewparseDeclaredSeverity(body)helper tolerant of markdown / punctuation shapes.buildCompiledRulehonors adeclaredSeverityOverrideoption with post-LLM deterministic override;severityOverridemarker threaded through rejection paths. Telemetry via newonSeverityOverridecallback recorded to.totem/temp/telemetry.jsonltaggedtype: 'severity-override'. Both compile prompts gain a Declared Severity directive section; every schema and few-shot example now carries"severity".LC impact
liquid-citytest-contract rules ("Normalize temp paths for cross-platform equality","Spy on logger contracts in tests") previously shipped with scopes that excluded tests and silently never fired. Retriage viatotem compile --upgrade <hash>closes this on a follow-up chore PR.liquid-cityADR-008 authoring on PR 77 burned ~10 manual severity-edit commits across R2 + R3 rounds because the compile pipeline emitted"severity": "warning"despite lesson prose declaringSeverity: error. The mechanical re-edit loop goes away.Scope
.changeset/1.15.4-lc-velocity-classifier-improvements.md— patch-level across the fixed group. Two-section release notes.docs/active_work.md— preamble rewritten for 1.15.4. Recently Shipped entry covers both PRs with SHAs + per-feature detail + LC-queue state + bot-review tail from feat(core): honor declared severity from lesson prose in compile output (#1656) #1658 (four rounds, all addressed)..strategysubmodule bump113179c→7892892b. Picks up strategy PR fix: LanceDB deleteByFile case-sensitivity + core test suite #125 (upstream-feedback items 015 + 016 from LC session-17) and strategy PR Epic: Frictionless 10-Minute Init (totem init) #124 (upstream-feedback item 017 — three-layer language-support-gap addendum for the pending Rust-support arc).No code changes in this PR; the two feature PRs already landed on main with their own test suites passing.
Test plan
pnpm run formatpnpm exec totem lintPASS (0 violations)pnpm exec totem verify-manifestPASS (443 rules)pnpm exec totem review(deterministic fast-path, non-code changes)🤖 Generated with Claude Code