Code review powered by the intelligence layer.
Codewalk's review engine understands your architecture, knows what files are risky, and reviews diffs with full context — not just syntax.
REVIEW_INTELLIGENCE
Built on the codebase intelligence layer
It doesn't just lint — it combines git diff, static analysis, graph risk, and team guidelines so the host LLM reviews each change with the same context a senior engineer would bring.
How it works
git diff → Static Analysis
graph risk · PageRank · cycles · blast radius
batch files (3-5 per batch, grouped by feature)
host LLM reviews each batch with full context
submit findings to disk per batch
final summary: all findings + verdict
Works with or without indexing
Review runs on any git repo — no analyze step needed. The intelligence layer enhances it when available.
Component
No index
With index
Git diff + content
Rubrics + guidelines
Stack detection
Risk annotations ⚠️ → ✅
Neighborhood context ⚠️ → ✅
Blast radius ⚠️ → ✅
Severity levels
blockerMust fix before merge
errorReal bugs, logic errors, security risks
suggestionStyle, naming, minor improvements
| Capability | Typical tools | Codewalk |
|---|---|---|
| Architecture awareness | No dependency graph | DuckDB + igraph: PageRank, fan-in, cycles, bottlenecks |
| Blast radius | — | "This file has 23 callers — review with extra care" |
| Works without indexing | — | Just needs a git repo (graph enhances but isn't required) |
| Batched for large PRs | Dumps everything at once | 3-5 files per batch, sorted by risk, host LLM stays focused |
| Custom rubrics | Limited | Per-language + per-framework + team guidelines |
| Fix application | Suggests only | Accept/reject → apply atomically → verify with tests |
Run a review from the API
POST /review gathers the diff, runs static analysis, batches files by risk, reviews with the configured LLM, and returns a structured verdict with findings.
curl -X POST http://localhost:8000/review \
-H "Content-Type: application/json" \
-d '{
"staged": false,
"target_branch": "master"
}'