Kb Blast Radius

Blast Radius Analysis

Lightweight, zero-dependency reverse dependency analysis using grep-based import tracing. Answers: "if I change file X, what else is affected?"

When to Use

• Planner: Before task decomposition — identify coupled files so tasks have correct boundaries and file ownership.
• Reviewer: After implementation — verify the implementer considered all dependents of changed files.
• Reviewer (scope: branch): During the cross-task integration sweep — structurally validate that all dependents were covered across the full changeset.

How It Works

1. Collects all source files in the repo (excluding node_modules, vendor, dist, build, target, .git)
2. Greps each file for import/require/use statements
3. Resolves relative imports to file paths
4. Builds a reverse dependency map: for each file, which files import it
5. BFS-expands from seed files through the reverse map (depth-limited)
6. Matches seed files and dependents to test files by naming convention

Supports: TypeScript, JavaScript, Python, Go, Rust, Java, Kotlin, Ruby, PHP, C/C++, Vue, Svelte.

Running the Script

The script is at ${CLAUDE_SKILL_DIR}/scripts/blast-radius.sh. Invoke it from the project's working directory.

Pre-implementation (planner)

When you know which files will be modified, pass them as targets:

bash ${CLAUDE_SKILL_DIR}/scripts/blast-radius.sh \
  --target src/auth/token.ts src/auth/session.ts

Post-implementation (reviewer)

Analyze actual changes against a base ref:

bash ${CLAUDE_SKILL_DIR}/scripts/blast-radius.sh \
  --changed main

Defaults to HEAD~1 if no base ref given. Also picks up unstaged and untracked files if git diff is empty.

Options

Interpreting Results

The output categorizes affected files:

• Seed files — the files being changed (input)
• Direct dependents — files that directly import a seed file. These are most likely to break or need updates.
• Transitive dependents — files that import a direct dependent. Lower risk but worth checking for cascading changes.
• Associated tests — test files matched by naming convention (.test.ts, .spec.ts, _test.go, test_*.py, etc.)

Limitations

• File-level, not function-level. If file A imports one function from file B, the whole file shows as a dependent. This is a conservative over-approximation — better to flag too many files than too few.
• Static imports only. Dynamic imports (import() with variables), reflection, and metaprogramming are not captured.
• No re-export tracing. If file A re-exports from file B, dependents of A won't appear as dependents of B.
• C# not supported. Namespace-based imports don't map to file paths without project analysis.
• Go is module-aware but only works if go.mod exists at the root.

These gaps mean the script covers ~80-90% of real dependency edges. When the output says "no dependents found," it may be a leaf node or may use a pattern the script doesn't capture — agents should use judgment.

For Planners

Use blast radius to inform task design:

1. Run --target on the files you intend to modify
2. If two seed files share many dependents, they likely belong in the same task
3. If a seed file has many direct dependents, consider whether dependents need API updates (breaking change) or are unaffected (internal refactor)
4. Include high-impact dependents in the task's "Files owned" list
5. Flag transitive dependents in the task's review checklist

For Reviewers

Use blast radius to verify implementation completeness:

1. Run --changed against the base branch
2. Check: were all direct dependents either updated or confirmed unaffected?
3. If a direct dependent wasn't touched and isn't in the task's scope, flag it — the implementer may have missed a required update
4. Verify associated test files were run or updated

For Reviewer (scope: branch)

Use blast radius for the cross-task integration sweep:

1. Run --changed against the base branch to see the full changeset's impact
2. Cross-reference with the plan's integration contracts
3. Any direct dependent that appears in the blast radius but wasn't touched by any task is a potential integration gap