| name | light-code-CheckOsCompatibility |
| description | Audit Delphi FMX source for cross-platform issues — Windows-only APIs, hardcoded paths, missing platform conditionals, mobile-incompatible patterns. Say "check this for Android/iOS", "is this safe on mobile/macOS/Linux", "audit for cross-platform issues". |
/light-code-CheckOsCompatibility — Cross-Platform Compatibility Audit
This is a thin launcher. Your only job is to resolve the input scope and launch the
light-code-CheckOsCompatibility agent. You do NOT audit the code yourself.
Step 1 — Resolve the input
The argument is in $args. The user may pass a single .pas file, a folder, or several files.
Build the audit set:
- One or more explicit file paths → use them as-is.
- A folder → Glob
<folder>/**/*.pas for the file list. If the result is large (more than
~25 files), tell the user the count and ask whether to audit all or narrow the scope.
- No argument → audit the
.pas files with uncommitted changes. Run
git status --porcelain and keep the .pas files it lists (modified, added, or renamed) —
this is "the code I just changed". If that list is empty, ask the user which file or folder to
audit. Do NOT default to the whole repository.
If the user names a target platform (Android, iOS, macOS, Linux), carry that into the agent's
prompt so it focuses there. Print the resolved audit set as a short list before launching.
Step 2 — Launch the agent
Call the Agent tool with subagent_type: "light-code-CheckOsCompatibility". Pass it the full audit set
in one prompt, plus any target platform the user mentioned. The agent reads the project
CLAUDE.md for the declared target platforms and intentional platform-specific code, scans the
files, and reports issues (Windows-only units/APIs, hardcoded paths, missing {$IFDEF} guards,
mobile-incompatible patterns).
Relay its final report to the user.
Step 3 — Summary
Print a short summary (the agent's full report is already in the transcript — do not re-paste
it): how many files audited, how many issues by severity, and which need a human decision.
Rules
- You do not audit code. Resolving the scope, launching the agent, and summarizing is your
entire job. The agent does the audit.
- This agent only reports (its tools are read + web). It does not apply fixes — so do not
promise fixes; relay the findings and let the user decide.
