From iris-dev
Automatically reviews generated ObjectScript code for common LLM mistakes including QUIT/RETURN misuse, method call syntax, error handling patterns, globals usage, and timestamp formats before presenting to the user.
How this skill is triggered — by the user, by Claude, or both
Slash command
/iris-dev:objectscript-reviewThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Automatically confirm generated ObjectScript follows the critical project rules before showing it to the user.
Automatically confirm generated ObjectScript follows the critical project rules before showing it to the user.
Do not show ObjectScript code to the user until this review passes.
For each item, check the generated code and flag any violations:
Quit <value> inside TRY/CATCH or loops..MethodName() syntax$$$ThrowOnError / $$$ISERR macros, not raw status checks%Status directly — uses %Exception objects$TLevel > 0 firstNew command inside method/procedure blocksYYYY-MM-DD HH:MM:SS format, not ISO 8601 with T$$$OK and check with $$$ISOK/$$$ISERRIf violations found:
⚠️ ObjectScript review flagged [N] issues — correcting before showing:
- [rule]: [what was wrong] → [correct pattern]
Then show the corrected code.
If clean:
✅ ObjectScript review passed.
Then show the code.
npx claudepluginhub intersystems-community/iris-agentic-devHard gate checklist that catches common ObjectScript mistakes before showing code to the user. Covers Quit/Return, postfix syntax, $IsObject checks, SQL table names, SQLCODE, HTML escaping, arithmetic, $ListBuild, %Status, transactions, storage blocks, %INLIST, and '= in SQL strings.
Provides checklists for code reviews assessing correctness, security vulnerabilities, performance, and maintainability. Use for pull requests, audits, and feedback.
Provides a structured code review checklist covering correctness, security, performance, and maintainability. Activates when reviewing pull requests, code audits, or identifying vulnerabilities.