detect_impact

Pre-commit change analysis with scope selection, affected processes, and risk assessment

Overview

The detect_impact prompt guides AI agents through a comprehensive pre-commit impact analysis workflow. It analyzes your current code changes, identifies affected execution flows, and provides a risk assessment.

Use this before committing to understand the blast radius of your changes.

Parameters

scope · string · default: "all"

What changes to analyze:

• unstaged — Only unstaged changes
• staged — Only staged changes
• all — All uncommitted changes (staged + unstaged)
• compare — Compare against a specific branch/commit (requires base_ref)

base_ref · string

Branch or commit to compare against (required if scope is compare)

Examples: main, develop, HEAD~1, a1b2c3d

What It Does

The prompt instructs the agent to:

1. Run detect_changes with the specified scope to find changed symbols and affected processes
2. Analyze critical symbols using context tool to see full reference graph
3. Check blast radius using impact tool for high-risk items
4. Generate risk report with changes, affected processes, risk level, and recommendations

Usage

In Claude Desktop

Run the detect_impact prompt

Agent will ask for parameters or use defaults.

In Cursor/OpenCode

Invoke the prompt from the MCP prompts menu, or reference it directly:

Use the detect_impact prompt to analyze my changes

With Parameters

Run detect_impact with scope="staged"

Run detect_impact comparing against main branch

Example Workflow

Here’s what the agent does when you invoke this prompt:

1. Detect Changes

   Runs detect_changes({scope: "all"}) to find modified symbols and affected processes

   changed_symbols:
     - name: validateUser
       file: src/auth/validate.ts
       change_type: modified
       affected_processes: ["User Login Flow", "Password Reset"]
   
2. Analyze Critical Symbols

   For each symbol in critical processes, runs context({name: "validateUser"}) to see:

   • All callers
   • All callees
   • Process participation
   • Related symbols
3. Check Blast Radius

   For high-risk symbols (many callers or cross-process usage), runs:

   impact({target: "validateUser", direction: "upstream"})
   

   To see what breaks at depth 1, 2, and 3.

4. Generate Report

   Summarizes findings:

   # Impact Analysis
   
   ## Changes
   - Modified `validateUser` in src/auth/validate.ts
   - Modified `hashPassword` in src/auth/utils.ts
   
   ## Affected Processes
   - User Login Flow (12 steps)
   - Password Reset (8 steps)
   - Admin Login (10 steps)
   
   ## Risk Level: HIGH
   
   `validateUser` has 23 upstream callers across 3 processes.
   Breaking change could affect all authentication flows.
   
   ## Recommendations
   1. Add tests for all 3 affected processes
   2. Review upstream callers for compatibility
   3. Consider deprecation path if changing signature
   

Example Output

The agent presents a structured risk report:

# Pre-Commit Impact Analysis

## Summary
Analyzing all uncommitted changes in repository "my-app"

## Changed Symbols
1. **validateUser** (src/auth/validate.ts)
   - Change: Modified function signature
   - Callers: 23 direct, 47 indirect
   - Processes: User Login Flow, Password Reset, Admin Login

2. **hashPassword** (src/auth/utils.ts)
   - Change: Updated hashing algorithm
   - Callers: 5 direct, 12 indirect
   - Processes: User Registration, Password Reset

## Affected Processes

### High-Risk Processes
- **User Login Flow** (12 steps)
  - Changed symbols: validateUser (step 2), hashPassword (step 3)
  - Risk: Breaking change would prevent all user logins

- **Password Reset** (8 steps)
  - Changed symbols: validateUser (step 4), hashPassword (step 6)
  - Risk: Could break password recovery

### Medium-Risk Processes
- **Admin Login** (10 steps)
  - Changed symbols: validateUser (step 3)
  - Risk: Admin access could be affected

## Blast Radius

### validateUser
- **Depth 1** (WILL BREAK): 23 functions
  - loginHandler, resetPasswordHandler, adminLoginHandler, ...
- **Depth 2** (MAY BREAK): 47 functions
- **Depth 3** (WATCH): 89 functions

### hashPassword
- **Depth 1** (WILL BREAK): 5 functions
- **Depth 2** (MAY BREAK): 12 functions

## Risk Assessment

**Overall Risk: HIGH**

- Core authentication functions modified
- 3 critical processes affected
- 70+ functions in blast radius
- No breaking changes detected in signatures

## Recommendations

1. **Test Coverage**
   - Add integration tests for all 3 affected processes
   - Verify backward compatibility with existing callers

2. **Review**
   - Manually verify the 23 direct callers of validateUser
   - Check if hashPassword algorithm change affects stored hashes

3. **Deployment**
   - Consider gradual rollout
   - Monitor authentication metrics post-deploy

4. **Documentation**
   - Update API docs if function signatures changed
   - Add migration guide if breaking

When to Use

Before Committing

Run this before every commit to catch breaking changes

Before PR

Include the report in your PR description

During Code Review

Use to understand reviewer’s concerns about blast radius

Before Deploy

Final check before merging to production

Comparison with Tools

This prompt orchestrates multiple tools:

Tool	Purpose in Workflow
detect_changes	Find what changed and affected processes
context	Get 360-degree view of critical symbols
impact	Calculate blast radius for high-risk items

The prompt provides the workflow logic while tools provide the data.

Next Steps

Detect Changes Tool

Learn about the underlying tool used by this prompt