Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →yeachan-heo/trace
Evidence-driven tracing lane that orchestrates competing tracer hypotheses in Claude built-in team mode
Passed 12 of 17 test cases with the skill, and 9 of 17 without it.
and 13 more in the full report
8 fixed8 passed either way2 still failing4 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 8 test cases the skill fixed. Same model, same prompt, with and without the skill.
During stress testing, our streaming event router randomly drops 5% of messages under high throughput, but logs show no socket errors or dropped packets at the network layer. Investigate why this message loss is occurring. Formulate candidate explanations, collect evidence pro and con for each, and conduct a rebuttal exchange between the top two theories to see if they converge or remain distinct. Return a ranked list of explanations, why weaker options were demoted, the key missing fact, and the highest-value next experiment.
| Checked | Without skill | With skill |
|---|---|---|
| The initial hypothesis lanes evaluate code-path defects, config/orchestration shifts, and measurement or assumption mismatches | Missed | Passed |
| The analysis presents a Ranked Hypotheses table containing explicit Confidence and Evidence Strength ratings | Missed | Passed |
| The final synthesis includes a dedicated Most Likely Explanation section | Missed | Passed |
The instructions your model reads when the skill loads.
Use this skill for ambiguous, causal, evidence-heavy questions where the goal is to explain why an observed result happened, not to jump directly into fixing or rewriting code.
This is the orchestration layer on top of the built-in tracer agent. The goal is to make tracing feel like a reusable OMC operating lane: restate the observation, generate competing explanations, gather evidence in parallel, rank the explanations, and propose the next probe that would collapse uncertainty fastest.
Use /oh-my-claudecode:trace when the problem is:
Examples:
Always preserve these distinctions:
Do not collapse into:
test cases passed, vs 9 / 17 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.