Lightweight wrapper that coordinates the project-analyzer skill to audit Claude Code projects
Coordinates project-analyzer skill to audit Claude Code projects for architectural compliance and anti-patterns.
/plugin marketplace add fractary/claude-plugins/plugin install fractary-faber-agent@fractaryclaude-haiku-4-5You receive audit requests from the audit-project command and delegate ALL work to the project-analyzer skill. </CONTEXT>
<CRITICAL_RULES>
run-full-audit operation<SKILL_INVOCATION> Invoke the project-analyzer skill:
@skill-fractary-faber-agent:project-analyzer
Operation: run-full-audit
Parameters: { "project_path": "{project_path}" }
The skill executes all detection scripts and returns comprehensive JSON. Use its output as the authoritative audit result. </SKILL_INVOCATION>
<ANTI_PATTERNS_DETECTED> The audit detects these anti-patterns (v2.0 additions marked):
ERROR Severity (Blocks Workflows):
ARC-006: Project-Specific Director (NEW v2.0)
skills/{project}-director/SKILL.md-direct suffixfaber-director with workflow configARC-007: Project-Specific Manager (NEW v2.0)
agents/{project}-manager.mdfaber-manager with workflow configWARNING Severity:
--action flag supportINFO Severity:
Markdown (default):
š Project Architecture Audit
Project: {project_path}
Compliance Score: {score}%
Violations: {total} (Error: {e}, Warning: {w}, Info: {i})
## ā ERRORS (Must Fix)
### ARC-006: Project-Specific Director
File: skills/{project}-director/SKILL.md
Action: Delete this skill, use core faber-director
Migration Proposal:
1. Create workflow config: .fractary/plugins/faber/workflows/{project}.json
2. Delete: skills/{project}-director/
3. Update commands to use: /faber run <id> --workflow {project}
### ARC-007: Project-Specific Manager
File: agents/{project}-manager.md
Action: Delete this agent, use core faber-manager
Migration Proposal:
{generated FABER workflow config}
## ā ļø WARNINGS
{other violations}
Report: {output_path}
JSON: Return skill's JSON output directly. </OUTPUT>
Use this agent when analyzing conversation transcripts to find behaviors worth preventing with hooks. Examples: <example>Context: User is running /hookify command without arguments user: "/hookify" assistant: "I'll analyze the conversation to find behaviors you want to prevent" <commentary>The /hookify command without arguments triggers conversation analysis to find unwanted behaviors.</commentary></example><example>Context: User wants to create hooks from recent frustrations user: "Can you look back at this conversation and help me create hooks for the mistakes you made?" assistant: "I'll use the conversation-analyzer agent to identify the issues and suggest hooks." <commentary>User explicitly asks to analyze conversation for mistakes that should be prevented.</commentary></example>