roblox-debugging-bugfix
/roblox-debugging-bugfix
Debug and fix Roblox bugs through Studio MCP. Use for repro steps, console errors, script search/read/edit flow, broken UI, failed gameplay systems, data issues, remotes, regressions, root-cause reports, and post-fix verification.
- Author
- AshExplained ↗
- Repository
- roblox-skills ↗
- Licence
- MIT
- Last repository change
- 01.07.2026
Assessment and security
- Files in this package
- 1
- Scripts
- 0
- HTML comments
- 0
- URLs in SKILL.md
- 0
No signals detected. Repository changed on 01.07.2026, without references to retired models or APIs.
MIT, open: use and modify while retaining author attribution.
What it does and when to use it
Debug and fix Roblox bugs through Studio MCP. Use for repro steps, console errors, script search/read/edit flow, broken UI, failed gameplay systems, data issues, remotes, regressions, root-cause reports, and post-fix verification.
Original author description
Debug and fix Roblox bugs through Studio MCP. Use for repro steps, console errors, script search/read/edit flow, broken UI, failed gameplay systems, data issues, remotes, regressions, root-cause reports, and post-fix verification.
How it works
- Goal
- Phase 1 — Build a repro loop first (this is the skill)
- Phase 2 — Reproduce
- Phase 3 — Hypothesize
- Phase 4 — Instrument
- Phase 5 — Fix + regression check
- Debug Loop (mechanics)
- Common Bug Classes
- Fix Rules
- Output
SKILL.md preview
---
name: roblox-debugging-bugfix
description: Debug and fix Roblox bugs through Studio MCP. Use for repro steps, console errors, script search/read/edit flow, broken UI, failed gameplay systems, data issues, remotes, regressions, root-cause reports, and post-fix verification.
---
# Roblox Debugging Bugfix
## Goal
Move from symptom to verified fix without guessing. For hard or intermittent bugs, follow the discipline below; skip phases only when you can justify it.
## Phase 1 — Build a repro loop first (this is the skill)
Everything else is mechanical once you have a fast, repeatable pass/fail signal for the bug. Spend disproportionate effort here. Be aggressive and creative before you touch a fix.
Ways to construct one, roughly in order:
1. **Studio playtest repro** — the exact steps in Play mode that surface the bug; watch `get_console_output`.
2. **`execute_luau` harness** — call the suspect server/module function directly with a fixture input and assert the result, instead of clicking through the whole game.
3. **Targeted scenario** — spawn the minimal instances/state that trigger the code path (one part, one remote, one player state).
4. **Replay a captured payload** — save the real remote payload / data profile that misbehaves and push it back through the code path in isolation.
Make the loop **fast, sharp, and deterministic**: narrow to the specific symptom, pin any randomness/timing, and cut unrelated setup. A 2-second deterministic check beats a 30-second flaky one.
**Intermittent bugs:** the goal is a higher reproduction rate, not a clean one-shot. Loop the trigger many times, add stress, narrow timing windows. A 50%-flake bug is debuggable; keep raising the rate until it is.
If you genuinely cannot build a loop, stop and say so — list what you tried and ask the user for the environment, a captured artifact (log dump, screen recording with timestamps, the offending data), or permission to add temporary instrumentation. Do not hypothesize without a loop.
## Phase 2 — Reproduce
Run the loop. Confirm it produces the failure the **user** described (not a nearby lookalike), that it repeats, and that you've captured the exact symptom for later verification.
## Phase 3 — Hypothesize
Generate **3–5 ranked, falsifiable hypotheses** before testing any — single-hypothesis debugging anchors on the first guess. Each states a prediction: "If X is the cause, then changing Y makes it disappear." Show the ranked list to the user before probing; they often re-rank instantly ("we just changed #3"). Proceed on your ranking if they're AFK.
## Phase 4 — Instrument
Each probe maps to one hypothesis; change one variable at a time. Prefer inspecting real values (`inspect_instance`, `execute_luau`) over log spam. **Tag every temporary log** with a unique prefix like `[DEBUG-a4f2]` so cleanup is a single `script_grep`. For performance regressions, measure first (timing/`os.clock`, MicroProfiler, instance counts) and bisect — logs are usually the wrong tool.
## Phase 5 — Fix + regression check
Make the smallest safe fix with `multi_edit`. Before declaring done, turn the repro into a durable check where a correct seam exists (an `execute_luau` assertion or a documented playtest step). If no good seam exists, that itself is a finding — note that the architecture is preventing the bug from being locked down.
## Debug Loop (mechanics)
1. Capture the user-visible symptom and expected behavior.
2. Reproduce in Play mode when possible.
3. Read `get_console_output`.
4. Search with `script_grep` and `script_search`.
5. Read relevant scripts with `script_read`.
6. Inspect relevant instances with `search_game_tree` and `inspect_instance`.
7. Make the smallest safe fix with `multi_edit`.
8. Re-run the repro and check console output again.
9. Remove all `[DEBUG-...]` instrumentation (`script_grep` the prefix).
## Common Bug Classes
- Client/server state mismatch.
- Remote payload trust issue.
- UI not updating after server state changes.
- Event connections duplicated or never disconnected.
- DataStore load/save race.
- Object path renamed or missing.
- Play mode DataModel mismatch.
- Race between character spawn, UI creation, and server initialization.
## Fix Rules
- Prefer root-cause fixes over defensive silence.
- Do not swallow errors without logging context.
- Add validation near trust boundaries.
- Keep unrelated refactors out of bug fixes.
- Verify no new console errors.
## Output
Return:
- Repro.
- Root cause (state which hypothesis proved correct).
- Files/scripts changed.
- Verification performed.
- Residual risks.
- Post-mortem: what would have prevented this bug? If the answer is architectural (no good test seam, tangled client/server boundary, hidden coupling), recommend `roblox-code-maintenance-refactor` or `roblox-luau-architecture` — after the fix is in, not before.
## Next
Verify the fix holds with `roblox-playtest-qa` (and `roblox-mobile-playtest` if it touched UI/mobile). If the post-mortem surfaced architectural debt, hand off to `roblox-luau-architecture` or `roblox-code-maintenance-refactor`. Return to `roblox-game-development-lifecycle` to resume the roadmap.
Versions
The database has no other skills named roblox-debugging-bugfix.
Identical content appears in 1 other GitHub folders. We show one copy with the highest score.
Database snapshot: 13.11.2026 · package version c636440219de (SHA-256 digest of the entire package)
Downloads in the last 30 days: 0 · ranking
Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.
Similar skills
- 90ro-engineer
Build, debug, review, and design Roblox/Luau gameplay systems. Use for project integration, remotes, persistence, lifecycle, architecture, profiling, or gameplay loops. - 100roblox-mcp
Connect an AI client to Roblox Studio through MCP. Use for connection setup, Studio tools, multiple Studio windows, playtesting, or combining MCP with Script Sync. - 100roblox-studio-mcp
Use when working in Roblox Studio through MCP: editing scripts or scenes, generating assets, playtesting, screenshots, or input. - 90roblox-studio-workflow
Use Roblox Studio MCP safely and productively from Claude. Use when connecting to Studio, inspecting a Roblox place, editing scripts or instances, inserting assets, generating models/materials, or playtesting through MCP tools. - 90roblox-studio-mcp
Use the Roblox Studio MCP server (Roblox_Studio) to inspect the open place, read scripts, run Luau, play test, read Output, simulate input and take screenshots. Use whenever Studio MCP tools are available, to verify changes in Studio, or when asked to test or debug in the game. - 100roblox-testing
Test, debug, or profile Roblox experiences. Use for Studio debugging, TestEZ, logging, connection cleanup, memory or network diagnostics, MicroProfiler, Scene Analysis, and Script Profiler.