fix: make embed command respect .codegraphrc.json model config#96
fix: make embed command respect .codegraphrc.json model config#96carlos-alm merged 3 commits intomainfrom
Conversation
Sets Xenova/bge-large-en-v1.5 (1024d) as the default embedding model for codegraph self-analysis, replacing the default minilm.
|
Claude finished @carlos-alm's task —— View job PR Review: Configure bge-large as default embedding modelTodo List:
Phase 1: Root Cause AnalysisProblem Being Solved:
Why This Approach: ❌ FUNDAMENTALLY FLAWED
Phase 2: Configuration Change Detection (🚨 CRITICAL ISSUES)SEVERITY: HIGH-RISK COSMETIC CHANGE
Phase 3: Test Integrity ✅
Phase 4: Code Quality
Root Cause AnalysisProblem Being Solved: Mismatch between intended default (better quality model) and actual default (minilm) Critical Concerns
Final Recommendation
Suggested Fix: Modify CLI to use |
Greptile SummaryThis PR configures Key changes:
Confidence Score: 5/5
Important Files Changed
Last reviewed commit: 77ffffc |
The CLI embed command hardcoded 'minilm' as the default model via Commander, ignoring .codegraphrc.json config entirely. Now the embed command reads config.embeddings.model as the default when no -m flag is passed. Also fixes DEFAULTS.embeddings.model from 'nomic-v1.5' to 'minilm' to match the actual fallback used by the embedder, and updates the models command to show the configured default.
|
Claude finished @carlos-alm's task —— View job PR Review: Configure bge-large as default embedding modelTodo List:
Phase 1: Root Cause Analysis ✅Problem Being Solved:
Why This Approach: ✅ CORRECT NOW
Phase 2: Configuration Change Detection ✅SEVERITY: LOW-RISK BENEFICIAL CHANGE
Phase 3: Test Integrity ✅
Phase 4: Code Quality ✅
Root Cause AnalysisProblem Being Solved: CLI embed command ignored configuration files entirely, causing confusion and preventing users from setting project-wide defaults Critical ConcernsNONE - This PR actually fixes a legitimate bug while improving default model quality. Final Recommendation
Note: The previous Claude review was analyzing only the first commit and missed that a follow-up commit actually implemented the CLI fix. This PR is now complete and functional. |
Change DEFAULT_MODEL in embedder.js from 'minilm' to 'nomic-v1.5' to match the intended default. Import DEFAULT_MODEL in cli.js instead of hardcoding strings. The embed command now resolves the model as: CLI flag > config.embeddings.model > DEFAULT_MODEL. Restores config.js DEFAULTS.embeddings.model to 'nomic-v1.5' (was incorrectly changed in prior commit).
|
Claude finished @carlos-alm's task —— View job PR Review: Configure bge-large as default embedding modelTodo List:
Phase 1: Root Cause Analysis ✅Problem Being Solved:
Why This Approach: ✅ JUSTIFIED
Phase 2: Configuration Change Detection ✅SEVERITY: LOW-RISK BENEFICIAL CHANGE
Phase 3: Test Integrity ✅No test files modified - This is actually GOOD:
Phase 4: Code Quality ✅Implementation Quality:
Technical Implementation:
Root Cause AnalysisProblem Being Solved: CLI embed command completely ignored configuration files, causing user confusion and preventing project-wide embedding model defaults Critical ConcernsNONE - This PR addresses legitimate functionality gaps and improves code quality. Final Recommendation
This is a textbook example of fixing broken functionality while improving system quality. The three-commit progression (config → CLI fix → default correction) shows careful iterative development. |
* docs: add community leverage features to backlog (IDs 97-99) Add three Leiden-powered features: weighted community labels (#97), hierarchical decomposition (#98), and community-aware impact scoring (#99). Mark ID 96 (README dep count) as done after graphology removal. Update ID 54 to reference Leiden instead of Louvain. * fix(backlog): renumber community items to IDs 100-102, fix Leiden text IDs 97-99 conflicted with existing entry #97 (unified multi-repo graph). Renumbered to 100-102. Also corrected "Leiden/Louvain" to "Leiden" in the section description for consistency with the rest of the PR. * docs: rename Tier 1e′ to Tier 1e.1 for ASCII compatibility * docs: fix premature Leiden references in backlog The section description referenced src/graph/algorithms/leiden/ which does not exist yet (pending PR #545). Item #96 was incorrectly marked DONE when graphology is still a live dependency. Updated section to be forward-looking and reverted #96 to BLOCKED on #545. * docs: add #545 dependency to community leverage items 100-102 * docs: add #545 dependency to item #54 for Leiden consistency * docs: mark #96 as SUPERSEDED instead of BLOCKED Once #545 merges, the README's "3 runtime dependencies" claim becomes correct automatically — no further action needed. BLOCKED implied someone still needs to act on the item after #545 lands.
Summary
embedcommand to readconfig.embeddings.modelfrom.codegraphrc.jsoninstead of hardcoding the default-mflag >config.embeddings.model>DEFAULT_MODELDEFAULT_MODELinembedder.jsfrom'minilm'to'nomic-v1.5'to match the intended defaultmodelscommand to show the configured default (not hardcoded).codegraphrc.jsonsettingbge-largeas the local embedding model for codegraph self-analysisTest plan
loadConfig('.')returnsembeddings.model: 'bge-large'from.codegraphrc.json