idea-mcp
Skill2 starsChanged 34 days agoApache-2.0View on GitHub
Use whenever the user says to use the IntelliJ IDEA MCP / JetBrains IDE MCP tools, or mentions idea_* tools, working inside an open IntelliJ IDEA project, or any task that should go through the IDE (reading/searching/editing code, building, running, VCS, refactoring, or debugging) instead of raw shell/file tools. This is the entry point that routes to the right specialized guidance.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- accso/ai-monitoringword for word, first seen Nov 8, 2026
Other versions1 project uses a version of this.
Word for word · 1
- accso/ai-monitoring★ 2100% same
Edited · 0
None found.
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository3 other files it gives its agents.
- AGENTS.mdAGENTS.md
- ai-context-workflowSkill
- benchmarkSkill
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 3 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 35
Read only the reference (s) needed for the current step - don't read all of them upfront.
- Very specific: line 42
Read the matching file under references/ in this skill before calling any tool it covers.
- Distinctive: line 47
If the task spans multiple phases (explore -> debug -> fix -> verify), read each piece of guidance in turn as you reach that phase, in the …
- Distinctive: line 49
Never fall back to shell/file tools for something an IDE MCP tool already covers - the IDE-native tool understands language semantics …
- Very specific: line 51
Always pass projectPath when known, to avoid ambiguous resolution across multiple open projects.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
IntelliJ IDEA MCP - Router
This is the entry point for all work done through the IntelliJ IDEA MCP server. It does not contain the detailed tool
guidance itself - it routes to a bundled reference file for the job at hand. Read the matching reference below before
calling any IDE MCP tool (note: despite the idea_* naming used loosely below and by users referring to "IDEA MCP
tools" generically, this session's live tool schema exposes bare names with no idea_ prefix — e.g. read_file,
not idea_read_file; confirmed live 2026-08-13, see references/tool-routing-contract.md for the corrected table).
Read references/tool-routing-contract.md FIRST, before any other reference in this skill. It is the binding,
single-page decision table for which tool to use per action (including the git-commit exception); the other reference
files below give supporting detail, but the contract is what's actually enforced (via the session's tool_ledger
table - see that file for how compliance is logged and queried, not just stated).
Core principle
IDEA MCP tools operate on the live IDE index (semantic, not text-based). Prefer them over shell/grep/cat whenever the project is open in the IDE. Use them for precise, targeted operations once you know (or can find) the file/symbol in question.
Pick the right guidance
| Task | Read |
| Find/understand code: symbol lookup, usages, file structure, project dependencies/modules, compile errors/warnings | references/explore.md |
| Change code: create/edit files, apply text edits, reformat, rename/extract/move/change-signature refactorings | references/edit.md |
| Build the project, run configurations or tests, use the IDE terminal, VCS status/diff, run notebooks, generate/run custom inspections | references/ops.md |
| Working inside an IntelliJ plugin project or the JetBrains platform monorepo itself (extension points, plugin.xml, threading/lock requirements) | references/plugin-dev.md |
| Need runtime evidence: breakpoints, stepping, stack/frame values, expression evaluation, confirming whether code is reached | references/debugger.md |
Read only the reference (s) needed for the current step - don't read all of them upfront. A single task often chains
several: e.g. read references/explore.md to locate the bug -> read references/debugger.md to confirm root cause ->
read references/edit.md to fix it -> read references/ops.md to build/test.
Workflow
- Classify the request using the table above.
- Read the matching file under
references/in this skill before calling any tool it covers. - Each reference file opens with its own Activation Gate: a check of which tools must actually be available in the
current MCP session (and, for
references/plugin-dev.md, whether the project even applies) before that guidance should be used. Respect it - if a gate fails, state the blocker explicitly instead of silently falling back to a different toolset or guessing tool names. - If the task spans multiple phases (explore -> debug -> fix -> verify), read each piece of guidance in turn as you reach that phase, in the order the work actually proceeds.
- Never fall back to shell/file tools for something an IDE MCP tool already covers - the IDE-native tool understands language semantics (imports, references, code style) that text tools don't.
- Always pass
projectPathwhen known, to avoid ambiguous resolution across multiple open projects.
When NOT to use this router
- Open-ended "what does this module/service do?" questions with no specific target - prefer whatever semantic/context search tools are available for broad exploration, and only switch to IDEA MCP once you have a concrete file/symbol target.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.