One connection. A real first result.
Add the server, sign in and approve, then verify your workspace connection. No key to copy for an OAuth-capable agent.
Build a cited research brief
Turn inspected public evidence into a decision-ready brief with traceable claims, counterexamples, actual coverage and clear next questions.
Read guideChoose whether the unit is a post, thread or claim
Separate the object you count from the evidence you read, so one long discussion cannot silently become many independent observations.
Read guideMatch research depth to a reversible decision
Choose a proportionate evidence standard for experiments and lasting commitments without treating a quick scan as approval for a costly change.
Read guideSeparate a causal question from a descriptive one
Turn an ambiguous request about impact into a clear distinction between observed discussion and an effect the available evidence cannot establish.
Read guideTurn an abstract concept into observable evidence
Define a research concept through concrete statements and counterexamples before reviewers apply it to public conversations.
Read guideDraw a boundary between adjacent research topics
Prevent a study of one problem from absorbing nearby complaints by recording inclusion rules, exclusions and an explicit boundary-case register.
Read guideSet a minimum context requirement for interpretation
Decide how much surrounding conversation is necessary before a short public statement can support a research claim.
Read guideDefine when a bounded evidence review should stop
Set a decision-oriented stopping rule before searching so an attractive emerging narrative cannot determine when the review ends.
Read guideSeparate discovery from confirmation
Keep hypotheses developed during open-ended reading apart from the evidence later used to evaluate those hypotheses.
Read guideSet an evidence threshold before making a recommendation
Agree on the evidence needed for a particular action so vivid examples do not silently lower the standard for approval.
Read guideMaintain a register of competing explanations
Track several plausible explanations together so early agreement does not erase alternatives that fit the same public evidence.
Read guideUse maximum variation to discover different mechanisms
Select contrasting cases deliberately when the goal is to uncover different explanations, not to estimate how common each explanation is.
Read guideSeek cases that challenge the emerging explanation
Investigate exceptions to an early research pattern before deciding whether the pattern is narrow, conditional or poorly supported.
Read guideApply a systematic selection rule to an ordered corpus
Use a reproducible interval rule for a bounded review while checking whether the ordering creates a hidden pattern in the selected material.
Read guideUse quotas without claiming population representation
Balance a qualitative review across known research categories while keeping arbitrary review allocations separate from real-world category shares.
Read guideChoose conversations as clusters before reading replies
Sample discussion containers deliberately and preserve within-thread dependence instead of treating every reply as an independent observation.
Read guideFollow public references without mistaking reach for breadth
Use linked discussions to extend a qualitative review while documenting seed dependence and avoiding claims that the resulting network is complete.
Read guideAudit what a convenient evidence set leaves out
Identify how an easily available corpus shapes a research conclusion before using it to recommend changes beyond the reviewed material.
Read guideLook for evidence lost through survivorship
Check whether a review favors material that remained visible or successful while excluding abandoned efforts that are harder to observe.
Read guideAccount for who chooses to speak publicly
Distinguish public expressions from the experiences of people who did not post, especially when complaints or enthusiasm attract participation.
Read guideReserve a held-out set for evaluating a coding rule
Protect a separate review set from iterative rule development so an apparent improvement is not merely better fitting the examples already inspected.
Read guideDistinguish an original account from a later retelling
Assign evidence roles to firsthand statements, summaries and commentary so a widely repeated story does not look independently corroborated.
Read guideSeparate firsthand experience from attributed experience
Keep authors’ own public accounts distinct from reports about friends, customers or unspecified others when assessing experience claims.
Read guideDistinguish expressed plans from observed actions
Review public statements of intention without presenting them as completed purchases, migrations or adoption outcomes. Record uncertainty explicitly.
Read guide