jetbrains-inspection

Run JetBrains IDE inspections through the local inspection plugin with route-safe lifecycle handling.

5|1|Updated May 5, 2026
One-click install
npx skills add https://github.com/cbusillo/codex-skills --skill jetbrains-inspection-cbusillo
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: jetbrains-inspection
Source: https://github.com/cbusillo/codex-skills/tree/main/skills/jetbrains-inspection
Command: npx skills add https://github.com/cbusillo/codex-skills --skill jetbrains-inspection-cbusillo

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve? JetBrains IDE inspections catch framework and language issues that tests miss, but driving them from an agent requires safe worktree routing, IDE lifecycle management, and reliable verdict interpretation. This Skill wraps the local inspection plugin HTTP API in a deterministic helper so agents can inspect the exact worktree, classify results as GREEN, RED, or UNKNOWN, and avoid unsafe direct HTTP calls. ## Core Features & Use Cases - Route-safe inspection helper: The jb-inspect.py script resolves the exact worktree route, opens and claims projects under a lease-bound lifecycle, runs inspections, and cleans up only helper-owned projects. - Readiness and closeout flows: Commands like agent-inspect and inspect-closeout produce compact verdict envelopes with retry policy, proof failures, and next actions for PR, push, and hand-off validation. - Repository preparation and configuration: Reads qualityGate.inspection settings from .github/github.json, supports multi-IDE lanes, structured Python project preparation, and durable preparation receipts. - Use Case: Before telling a user their Kotlin or Python change is ready to merge, run the helper against the current worktree to confirm a GREEN inspection verdict or surface actionable IDE findings with file and line details. ## Quick Start Ask the agent to run the JetBrains inspection helper with agent-inspect against the current repository using the changed_files scope and report the verdict.

Frequently Asked Questions about jetbrains-inspection

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run JetBrains IDE inspections from an AI coding agent?

Run the jb-inspect.py helper with uv using the agent-inspect command and a scope such as changed_files. The helper resolves the exact worktree route, runs the inspection through the local plugin, and returns a compact verdict envelope.

How do I check code is ready to push with IDE inspections?

Use the inspect-closeout command before declaring work ready, safe to push, or safe to merge. It opens the exact worktree if needed, inspects it, fetches problems, and cleans up helper-opened projects with explicit cleanup status.

Which JetBrains IDEs does the inspection helper support?

The helper supports IntelliJ IDEA, PyCharm, and WebStorm, including stable and explicitly selected EAP installs. Repositories declare their preferred IDE in .github/github.json so lifecycle opens target the correct product.

Why does my inspection return UNKNOWN instead of GREEN?

UNKNOWN means the inspection could not prove a clean or red result, due to stale results, indexing, session drift, ambiguous routes, or missing IDE configuration. Report the verdict reason and next action rather than treating it as no problems found.

Can I call the JetBrains inspection plugin HTTP API directly with curl?

No, direct HTTP calls bypass route resolution, worktree safety, stale-result handling, lifecycle locking, and cleanup. The skill's command policy blocks such calls and requires the jb-inspect.py helper instead.

What are the limitations of worktree inspection routing?

Inspections must target the exact worktree being edited; a containing main checkout is insufficient for readiness. Auto-open only works under configured trusted roots, and the helper never closes preexisting or foreign-owned projects.