steering-matcher
Match file paths against steering file glob patterns to determine applicable steering guidance. Use when orchestrator needs to inject context-aware guidance based on files being modified.
When & Why to Use This Skill
The Steering File Matcher is a specialized Claude skill designed to optimize agent performance through dynamic, context-aware guidance injection. By matching file paths against glob patterns, it ensures that only the most relevant steering instructions are included in the agent's prompt, significantly reducing token usage by over 30% while improving the accuracy and relevance of the AI's output.
Use Cases
- Dynamic Context Injection: Automatically identify and inject specific coding standards, security protocols, or architectural guidelines based on the exact files an agent is currently modifying.
- Token Usage Optimization: Minimize LLM overhead and operational costs by filtering out irrelevant documentation, ensuring the context window remains focused on essential task-specific information.
- Multi-Module Workflow Orchestration: Manage complex projects where different directories require distinct rules, such as applying backend validation logic to API folders while using UI design systems for frontend components.
- Priority-Based Guidance Resolution: Handle overlapping instructions by sorting applicable steering files by priority, ensuring that critical global rules or specific project overrides are correctly applied.
| name | steering-matcher |
|---|---|
| description | Match file paths against steering file glob patterns to determine applicable steering guidance. Use when orchestrator needs to inject context-aware guidance based on files being modified. |
| license | MIT |
| version | 1.0.0 |
| model | claude-haiku-4-5 |
Steering File Matcher Skill
Purpose
This skill helps orchestrator determine which steering files to inject into agent context based on the files being modified.
- Match file paths against glob patterns in steering file front matter
- Return applicable steering files sorted by priority
- Enable token-efficient context injection (30%+ savings)
When to Use
Invoke this skill when you need to:
- Determine which steering guidance applies to a task
- Inject context-aware guidance into agent prompts
- Optimize token usage by scoping guidance to relevant files
- Match files against
applyTopatterns in steering files
Quick Usage
Using Get-ApplicableSteering.ps1 Script
The script in .claude/skills/steering-matcher/Get-ApplicableSteering.ps1 handles pattern matching.
# Match files against steering patterns
$files = @(
"src/claude/analyst.md",
".agents/security/TM-001-auth-flow.md"
)
$steering = pwsh .claude/skills/steering-matcher/Get-ApplicableSteering.ps1 `
-Files $files `
-SteeringPath ".agents/steering"
# Output: Array of hashtables with Name, Path, ApplyTo, Priority
foreach ($s in $steering) {
Write-Host "Matched: $($s.Name) (Priority: $($s.Priority))"
}
Workflow Integration
This skill integrates with the orchestrator workflow:
- Task Analysis: Orchestrator identifies files affected by task
- Pattern Matching: Use this skill to find applicable steering
- Context Injection: Include matched steering in agent prompt
- Token Optimization: Only inject relevant guidance
Standard Orchestrator Workflow
# 1. Identify files from task
$filesAffected = @(
"src/claude/security.md",
".agents/security/SR-001-oauth-review.md"
)
# 2. Get applicable steering
$applicableSteering = pwsh .claude/skills/steering-matcher/Get-ApplicableSteering.ps1 `
-Files $filesAffected
# 3. Inject into agent context
# "Relevant steering: agent-prompts.md (Priority 9), security-practices.md (Priority 10)"
Implementation
See Get-ApplicableSteering.ps1 for the PowerShell implementation.
Testing
Run Pester tests to verify pattern matching:
Invoke-Pester .claude/skills/steering-matcher/tests/Get-ApplicableSteering.Tests.ps1