\r \r

Debug System\r

\r Find, explain, and fix errors by identifying the root cause.\r \r Use this skill when:\r

• something is not working\r
• an error message appears\r
• a workflow fails\r
• an API request breaks\r
• behavior does not match expectations\r
• the same issue keeps happening\r \r This is an instruction-only skill.\r It does not execute code or access external systems unless the user provides the data.\r \r ---\r \r

What this skill does\r

\r This skill helps an agent:\r \r

• analyze errors and unexpected behavior\r
• identify the real root cause (not just symptoms)\r
• explain what went wrong in simple terms\r
• suggest concrete fixes\r
• guide step-by-step debugging\r
• prevent the issue from happening again\r \r ---\r \r

What this skill does not do\r

\r This skill does not:\r \r

• guess without evidence\r
• invent missing logs or results\r
• claim something is fixed without validation\r
• access external systems without user input\r
• provide fake certainty\r \r ---\r \r

Inputs\r

\r Use what the user provides:\r \r

• error messages\r
• logs\r
• code snippets\r
• API responses\r
• workflow steps\r
• expected vs actual behavior\r
• screenshots or descriptions\r \r If information is missing, state assumptions clearly.\r \r ---\r \r

Debugging process\r

\r

Step 1 — Understand the problem\r

Identify:\r

• expected behavior\r
• actual behavior\r
• when the issue occurs\r
• what changed recently\r \r ---\r \r

Step 2 — Analyze the error\r

\r Look for:\r

• exact error message\r
• stack trace (if available)\r
• failing step or node\r
• invalid inputs\r
• missing fields\r
• wrong data types\r
• permission/auth issues\r
• network/API failures\r \r ---\r \r

Step 3 — Find root cause\r

\r Do not stop at symptoms.\r \r Ask:\r

• what caused this error?\r
• what condition triggered it?\r
• is this reproducible?\r
• is it a config issue, logic issue, or external failure?\r \r ---\r \r

Step 4 — Propose fixes\r

\r For each issue, provide:\r \r

• root cause\r
• why it happens\r
• exact fix\r
• alternative fix (if needed)\r
• how to verify the fix\r \r ---\r \r

Step 5 — Prevent recurrence\r

\r Suggest:\r

• validation checks\r
• better error handling\r
• logging improvements\r
• retry mechanisms\r
• safer defaults\r \r ---\r \r

Output format\r

\r

Debug Report\r

\r

Problem Summary\r

• Expected:\r
• Actual:\r
• Context:\r \r

Root Cause\r

• ...\r \r

Evidence\r

• ...\r \r

Fix\r

\r

Recommended Fix\r

• ...\r \r

Alternative Fix\r

• ...\r \r

Verification Steps\r

1. ...\r
2. ...\r
3. ...\r \r

Prevention\r

• ...\r
• ...\r \r

Open Questions\r

• ...\r
• ...\r \r ---\r \r

Behavior rules\r

\r

• Focus on root cause, not just symptoms\r
• Be clear and actionable\r
• Do not assume missing data without stating it\r
• Prefer simple fixes over complex ones\r
• Explain in a way the user can follow step-by-step\r \r ---\r \r

Quality checklist\r

\r Before finishing, ensure:\r \r

• root cause is identified\r
• fix is actionable\r
• verification steps are clear\r
• no fake assumptions\r
• prevention is included\r \r ---\r \r

When not to use this skill\r

\r Do not use this skill for:\r

• general explanations without a problem\r
• theoretical discussions without real issues\r
• tasks unrelated to debugging\r \r Use only when there is a failure, bug, or unexpected behavior.