Orchestrate evidence-driven tracing to evaluate competing hypotheses for ambiguous results.
npx skills add https://github.com/Cheggin/request-for-startups --skill trace-cheggin
Please help me install this Agent Skill. Skill: trace Source: https://github.com/Cheggin/request-for-startups/tree/main/skills/trace Command: npx skills add https://github.com/Cheggin/request-for-startups --skill trace-cheggin
Trace orchestrates evidence-driven reasoning to resolve ambiguous results by organizing competing hypotheses, gathering supporting and conflicting evidence, and ranking explanations to identify the most plausible root cause.
Provide a concrete observation and initiate the team-mode tracing workflow.
High-intent search queries and answers about installing and using this skill.
Debug ambiguous issues by restating observations and generating multiple competing hypotheses. Evaluate each in parallel tracer lanes, gathering supporting and conflicting evidence, ranking explanations, and running rebuttal rounds to converge on the most plausible root cause.
Yes, team mode tracing applies to performance regressions, architecture reviews, and experimental results. The workflow evaluates multiple explanations in parallel using three default tracer lanes to weigh evidence for and against each hypothesis.
Evaluate competing hypotheses by spawning parallel tracer lanes that gather evidence for and against each explanation. Rank evidence strength, identify critical unknowns, propose discriminating probes, and run rebuttal rounds to test leading explanations before synthesis.
The best way to weigh multiple explanations is evidence-driven reasoning that generates deliberately different hypotheses. Gather evidence for and against each across parallel lanes, rank explanations by evidence strength, and run rebuttals to surface convergence or separation before synthesizing.
Yes, evidence tracing workflows require Claude built-in team mode to spawn three default tracer lanes. Each lane independently pursues evidence for and against its assigned hypothesis, enabling parallel evaluation, ranking, and rebuttal rounds before synthesis.
Identify the critical unknown by gathering and ranking evidence strength across competing hypotheses. Once evidence is ranked, propose the best discriminating probe that separates leading explanations, then run a rebuttal round to test convergence before final synthesis.
oneforceOrchestrate parallel tracer hypotheses to identify evidence-based causal explanations.
cheafiGenerate competing causal hypotheses and validate them against evidence.
OliverOuyangOrchestrate parallel hypothesis tracing and evidence collection for causal failure analysis.
ThierryNExecute parallel debugging sessions with three agents investigating competing hypotheses.
