[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"org-braintrust":3,"repo-skills-v-0-3-0":110},{"org":4,"repos":53},{"slug":5,"name":6,"logoUrl":7,"githubOrg":8,"website":9,"skillCount":10,"repoCount":11,"topRepos":12,"topTags":23,"lastUpdatedAt":52},"braintrust","Braintrust","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fbraintrust.png","braintrustdata","https:\u002F\u002Fwww.braintrust.dev",27,4,[13,16,19,21],{"name":14,"skillCount":15},"braintrustdata\u002Feval-library",24,{"name":17,"skillCount":18},"braintrustdata\u002Fbraintrust-claude-plugin",1,{"name":20,"skillCount":18},"braintrustdata\u002Fbraintrust-codex-plugin",{"name":22,"skillCount":18},"braintrustdata\u002Fbraintrust-skills",[24,25,28,31,34,37,40,43,46,49],{"slug":5,"name":6},{"slug":26,"name":27},"evals","Evals",{"slug":29,"name":30},"llm","LLM",{"slug":32,"name":33},"agents","Agents",{"slug":35,"name":36},"analysis","Analysis",{"slug":38,"name":39},"datasets","Datasets",{"slug":41,"name":42},"debugging","Debugging",{"slug":44,"name":45},"documentation","Documentation",{"slug":47,"name":48},"strategy","Strategy",{"slug":50,"name":51},"experiments","Experiments","2026-08-20T03:53:40.036077",[54,69,84,100],{"name":55,"fullName":14,"repoUrl":56,"skillCount":15,"stars":57,"forks":58,"description":59,"topics":60,"topTags":61,"topTagCount":68,"lastUpdatedAt":52},"eval-library","https:\u002F\u002Fgithub.com\u002Fbraintrustdata\u002Feval-library",7,0,"Braintrust eval skills library",[],[62,63,64,65,66,67],{"slug":5,"name":6},{"slug":26,"name":27},{"slug":29,"name":30},{"slug":32,"name":33},{"slug":35,"name":36},{"slug":38,"name":39},29,{"name":70,"fullName":17,"repoUrl":71,"skillCount":18,"stars":72,"forks":73,"description":74,"topics":75,"topTags":76,"topTagCount":82,"lastUpdatedAt":83},"braintrust-claude-plugin","https:\u002F\u002Fgithub.com\u002Fbraintrustdata\u002Fbraintrust-claude-plugin",18,11,null,[],[77,78,79],{"slug":5,"name":6},{"slug":41,"name":42},{"slug":80,"name":81},"mcp","MCP",3,"2026-07-12T08:36:13.889274",{"name":85,"fullName":20,"repoUrl":86,"skillCount":18,"stars":87,"forks":58,"description":88,"topics":89,"topTags":90,"topTagCount":98,"lastUpdatedAt":99},"braintrust-codex-plugin","https:\u002F\u002Fgithub.com\u002Fbraintrustdata\u002Fbraintrust-codex-plugin",2,"Monorepo for Braintrust Codex Plugins",[],[91,94,95,96,97],{"slug":92,"name":93},"analytics","Analytics",{"slug":5,"name":6},{"slug":44,"name":45},{"slug":26,"name":27},{"slug":29,"name":30},5,"2026-07-12T08:36:17.655627",{"name":101,"fullName":22,"repoUrl":102,"skillCount":18,"stars":18,"forks":58,"description":74,"topics":103,"topTags":104,"topTagCount":11,"lastUpdatedAt":109},"braintrust-skills","https:\u002F\u002Fgithub.com\u002Fbraintrustdata\u002Fbraintrust-skills",[],[105,106,107,108],{"slug":32,"name":33},{"slug":5,"name":6},{"slug":26,"name":27},{"slug":29,"name":30},"2026-07-12T08:36:16.200552",{"items":111,"total":15},[112,127,140,152,166,180,192,204,220,232,244,258,272,286,298,308,323,333,347,359,373,383,397,413],{"slug":113,"name":113,"fn":114,"description":115,"org":116,"tags":117,"stars":57,"repoUrl":56,"updatedAt":126},"braintrust-analyze-eval-experiment","analyze LLM and agent eval experiments","Analyze completed LLM or agent eval experiments using uncertainty-aware and decision-relevant methods. Use to audit run completeness and pairing, calculate confidence intervals, run paired comparisons, report wins, losses, and ties, incorporate run-to-run variance, handle multiple comparisons, inspect subgroup performance, and test fragility to favorable slices. Use when results already exist and someone asks what they mean, whether a difference is real, or which model won. Do not use to design an experiment that has not yet collected results.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[118,120,121,122,123],{"name":36,"slug":35,"type":119},"tag",{"name":6,"slug":5,"type":119},{"name":27,"slug":26,"type":119},{"name":30,"slug":29,"type":119},{"name":124,"slug":125,"type":119},"Statistics","statistics","2026-08-20T03:53:01.13806",{"slug":128,"name":128,"fn":129,"description":130,"org":131,"tags":132,"stars":57,"repoUrl":56,"updatedAt":52},"braintrust-attribute-multi-variable-change","attribute performance changes to multiple variables","Attribute an observed change when several things moved at once — model plus prompt plus tools, a provider migration, a framework upgrade, or a vendor swap that bundles serving stack with model. Use when asked which part of a change caused the result, when a comparison's arms differ in more than one way, when a treatment has no uniform implementation across vendors, or when a serving-stack difference is confounded with a model difference. Do not use for a clean single-variable comparison, or to design an experiment that has not yet run.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[133,134,135,136,137],{"name":36,"slug":35,"type":119},{"name":6,"slug":5,"type":119},{"name":27,"slug":26,"type":119},{"name":30,"slug":29,"type":119},{"name":138,"slug":139,"type":119},"Performance","performance",{"slug":141,"name":141,"fn":142,"description":143,"org":144,"tags":145,"stars":57,"repoUrl":56,"updatedAt":151},"braintrust-build-eval-dataset","create and manage LLM eval datasets","Create, edit, audit, or compare eval datasets for LLM applications and agents, including target-population definition, case sourcing from production traces, stratified sampling, label provenance and label audits, expected values as constraints for open-ended tasks, dev\u002Ftest splits, contamination and leakage controls, headroom checks, refresh policy, and datasheets. Use when working on the content or lifecycle of an eval dataset. Do not use for sample-size or power calculations, scorer implementation, or open-ended adversarial failure discovery.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[146,147,148,149,150],{"name":33,"slug":32,"type":119},{"name":6,"slug":5,"type":119},{"name":39,"slug":38,"type":119},{"name":27,"slug":26,"type":119},{"name":30,"slug":29,"type":119},"2026-08-20T03:53:33.304815",{"slug":153,"name":153,"fn":154,"description":155,"org":156,"tags":157,"stars":57,"repoUrl":56,"updatedAt":165},"braintrust-define-eval-objective","define LLM evaluation objectives","Create, edit, or audit an eval objective by working backward from a product decision to the target outcome, construct, population, intended claim, and verification-versus-validation questions. Use when a team is unsure what an eval should establish, asks \"what are we actually trying to measure,\" \"is this eval measuring the right thing,\" \"does this benchmark support our claim,\" or needs to turn a product goal into an eval objective and state which claims are out of scope. Do not use to select detailed metrics, design datasets, or implement scorers.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[158,159,160,161,164],{"name":6,"slug":5,"type":119},{"name":27,"slug":26,"type":119},{"name":30,"slug":29,"type":119},{"name":162,"slug":163,"type":119},"Product Management","product-management",{"name":48,"slug":47,"type":119},"2026-08-20T03:53:00.07097",{"slug":167,"name":167,"fn":168,"description":169,"org":170,"tags":171,"stars":57,"repoUrl":56,"updatedAt":179},"braintrust-define-eval-release-gate","configure release gates for LLM applications","Create, edit, audit, or apply release gates for LLM applications and agents. Use to combine minimum meaningful improvement, statistical significance, regression rate, subgroup consistency, worst-run stability, all-attempts reliability, safety upper bounds, latency, and cost into an explicit ship-or-hold policy, to turn metrics into a CI gate, or to explain why a candidate failed a gate and what evidence would justify reconsideration. Do not use for general result analysis without a deployment decision.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[172,173,174,177,178],{"name":33,"slug":32,"type":119},{"name":6,"slug":5,"type":119},{"name":175,"slug":176,"type":119},"CI\u002FCD","ci-cd",{"name":27,"slug":26,"type":119},{"name":30,"slug":29,"type":119},"2026-08-20T03:53:15.686158",{"slug":181,"name":181,"fn":182,"description":183,"org":184,"tags":185,"stars":57,"repoUrl":56,"updatedAt":191},"braintrust-deploy-evaluator","deploy evaluators to Braintrust","Take a validated scorer or classifier from definition to running instrument in Braintrust — scope selection, inline testing before saving, saving as an evaluator, attaching an online-scoring rule, activating it for new traffic, and backfilling history with a rewind. Use when a scorer needs to actually run against production logs, when an online-scoring rule needs to be created or changed, or when historical traces need scoring. Do not use to decide what the scorer should measure, to write its rubric, or to establish that it agrees with human judgment.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[186,187,190],{"name":6,"slug":5,"type":119},{"name":188,"slug":189,"type":119},"Deployment","deployment",{"name":27,"slug":26,"type":119},"2026-08-20T03:53:32.558937",{"slug":193,"name":193,"fn":194,"description":195,"org":196,"tags":197,"stars":57,"repoUrl":56,"updatedAt":203},"braintrust-design-eval-experiment","design controlled LLM eval experiments","Design or audit controlled eval experiments for model, prompt, retrieval, tool, guardrail, or agent-architecture changes. Use before data collection to state directional and minimum-effect hypotheses, name independent, dependent, and control variables including the serving environment and tool surface, choose paired designs, set repetitions and allocation, distinguish exploratory from confirmatory comparisons, and pre-specify stopping, exclusion, multiplicity, and analysis rules. Do not use primarily to analyze results already collected.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[198,199,200,201,202],{"name":6,"slug":5,"type":119},{"name":27,"slug":26,"type":119},{"name":51,"slug":50,"type":119},{"name":30,"slug":29,"type":119},{"name":48,"slug":47,"type":119},"2026-08-20T03:53:36.534554",{"slug":205,"name":205,"fn":206,"description":207,"org":208,"tags":209,"stars":57,"repoUrl":56,"updatedAt":219},"braintrust-design-eval-instrumentation","design trace and evaluation dataset schemas","Design the trace and eval-dataset schema for an LLM app or agent, and wire the system to emit it. Use when deciding what to log, designing a trace schema, setting up tracing or observability before evals, or when failures cannot be debugged or sliced from existing traces — covering inputs, outputs, spans for tool and LLM calls, state changes, metadata, resolved configuration, serving path, tool manifest, per-item status, attachments, and subgroup variables. Do not use to decide what the evidence should mean, or to write scorers.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[210,211,212,213,216],{"name":6,"slug":5,"type":119},{"name":39,"slug":38,"type":119},{"name":27,"slug":26,"type":119},{"name":214,"slug":215,"type":119},"Observability","observability",{"name":217,"slug":218,"type":119},"Tracing","tracing","2026-08-20T03:53:37.274703",{"slug":221,"name":221,"fn":222,"description":223,"org":224,"tags":225,"stars":57,"repoUrl":56,"updatedAt":231},"braintrust-design-eval-metric-bundle","create multi-objective evaluation metric bundles","Create, edit, audit, or compare a multi-objective eval metric bundle covering product quality, safety, reliability, latency, and cost. Use to choose metrics for an eval, define a goodness bundle, distinguish optimization metrics from non-regression guardrails, expose tradeoffs, audit a KPI or single composite score for Goodhart and metric-gaming risk, or answer \"what should improve and what must not regress.\" Do not use to design trace schemas, build datasets, or implement scoring methods.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[226,227,228,229,230],{"name":33,"slug":32,"type":119},{"name":6,"slug":5,"type":119},{"name":27,"slug":26,"type":119},{"name":30,"slug":29,"type":119},{"name":138,"slug":139,"type":119},"2026-08-20T03:53:16.078609",{"slug":233,"name":233,"fn":234,"description":235,"org":236,"tags":237,"stars":57,"repoUrl":56,"updatedAt":243},"braintrust-design-human-eval-review","design human evaluation and review workflows","Design or audit human evaluation workflows and golden datasets for LLM applications and agents. Use to set up expert review, select review cases, write reviewer instructions, assign raters, capture rationales and confidence, measure inter-rater agreement with kappa or alpha, adjudicate disagreements, and preserve reviewed examples with provenance as a versioned reference set. Do not use to elicit the criteria or rubric in the first place, to validate a scorer once reference labels exist, or to implement the scorer.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[238,239,240,241,242],{"name":33,"slug":32,"type":119},{"name":6,"slug":5,"type":119},{"name":39,"slug":38,"type":119},{"name":27,"slug":26,"type":119},{"name":30,"slug":29,"type":119},"2026-08-20T03:53:36.900129",{"slug":245,"name":245,"fn":246,"description":247,"org":248,"tags":249,"stars":57,"repoUrl":56,"updatedAt":257},"braintrust-discover-agent-failures","identify and classify agent failure modes","Search open-endedly for unanticipated agent failure modes and convert them into a named taxonomy and durable regression items. Use for requests to find out what goes wrong, surface unknown or silent failures, do error analysis over traces, cluster and triage production failures, or build a failure taxonomy — where the goal is discovering modes nobody thought to test rather than measuring a predefined criterion. Produces datasets and taxonomies, not headline scores. Do not use for adversarial attacks against a threat model, or to measure a known criterion.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[250,251,252,253,254],{"name":33,"slug":32,"type":119},{"name":6,"slug":5,"type":119},{"name":42,"slug":41,"type":119},{"name":27,"slug":26,"type":119},{"name":255,"slug":256,"type":119},"Triage","triage","2026-08-20T03:53:00.423941",{"slug":259,"name":259,"fn":260,"description":261,"org":262,"tags":263,"stars":57,"repoUrl":56,"updatedAt":271},"braintrust-discover-trace-topics","deploy trace topic discovery pipelines","Design, evaluate, and deploy a Topics pipeline that discovers clusters in trace traffic — preprocessor, facet prompt, no-match policy, clustering, and the automation that runs it. Use when the label set is unknown and has to come out of the data: finding what a product is actually used for, surfacing recurring issues, or building the categorical vocabulary a classifier will later enforce. Also use to evaluate or repair an existing facet or preprocessor. Do not use when the label set is already known and stable — that is a classifier — or to run a hypothesis-driven or adversarial failure hunt.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[264,265,268,269,270],{"name":6,"slug":5,"type":119},{"name":266,"slug":267,"type":119},"Data Pipeline","data-pipeline",{"name":27,"slug":26,"type":119},{"name":30,"slug":29,"type":119},{"name":217,"slug":218,"type":119},"2026-08-20T03:53:00.778959",{"slug":273,"name":273,"fn":274,"description":275,"org":276,"tags":277,"stars":57,"repoUrl":56,"updatedAt":285},"braintrust-elicit-eval-criteria","extract and define evaluation criteria","Extract evaluation criteria out of domain experts and real user desires, and capture them as reusable evaluation assets before any labeling or scoring begins — construct facets, anchored exemplars, adversarial traps, scoring guidance, audit rules, and the signals that reveal what users actually want. Use when nobody can say what \"good\" means, when a rubric does not exist yet, when expert knowledge lives only in reviewers' heads, or when validating that an eval reflects user desires rather than team assumptions. Do not use to run the labeling workflow or to compare a scorer against finished labels.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[278,279,280,283,284],{"name":6,"slug":5,"type":119},{"name":27,"slug":26,"type":119},{"name":281,"slug":282,"type":119},"Ideation","ideation",{"name":30,"slug":29,"type":119},{"name":48,"slug":47,"type":119},"2026-08-20T03:53:36.164477",{"slug":287,"name":287,"fn":288,"description":289,"org":290,"tags":291,"stars":57,"repoUrl":56,"updatedAt":297},"braintrust-eval","execute Braintrust evaluation experiments","Execute an eval end to end against a live Braintrust project — org and credential selection, finding and importing a dataset, writing the task and scorer code, running the experiment behind smoke gates and quota preflight, and tracing agentic `claude -p` runs. Use when an eval has to actually run: \"run this eval,\" \"set up an eval for X,\" \"compare these models in Braintrust.\" Do not use to design a dataset, scorer, experiment, or release gate in the abstract — the lifecycle cards own what to build and why; this owns making it happen.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[292,293,294],{"name":6,"slug":5,"type":119},{"name":27,"slug":26,"type":119},{"name":295,"slug":296,"type":119},"Testing","testing","2026-08-20T03:53:29.405249",{"slug":299,"name":299,"fn":300,"description":301,"org":302,"tags":303,"stars":57,"repoUrl":56,"updatedAt":307},"braintrust-map-eval-evidence","create and audit evaluation evidence maps","Create or audit an evidence map that connects eval constructs or behavior specifications to observable success signals, failure signals, proxy limitations, and possible gaming paths. Use when a quality such as helpfulness, safety, trust, correctness, or task success is not directly measurable, when asking \"how would we observe this,\" \"what evidence would show this behavior happened,\" or when a BEHAVIOR.md must be translated into trace-review signals. Do not use to define storage or schema fields, choose product metrics, or write scorers.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[304,305,306],{"name":36,"slug":35,"type":119},{"name":6,"slug":5,"type":119},{"name":27,"slug":26,"type":119},"2026-08-20T03:52:56.298223",{"slug":309,"name":309,"fn":310,"description":311,"org":312,"tags":313,"stars":57,"repoUrl":56,"updatedAt":322},"braintrust-monitor-production-evals","monitor production LLM and agent evaluations","Design or audit online evaluation of live LLM and agent traffic: trace sampling, online scoring coverage, alert thresholds and ownership, drift and failure-slice monitoring, incident review, and the pipeline that routes production failures back into the offline dataset. Use for questions about scoring production traces, monitoring quality after launch, catching regressions in the wild, alert fatigue, scorer drift, or closing the loop from incident to eval item. Do not use to design offline experiments or to interpret a controlled comparison.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[314,315,316,317,318,321],{"name":33,"slug":32,"type":119},{"name":6,"slug":5,"type":119},{"name":27,"slug":26,"type":119},{"name":30,"slug":29,"type":119},{"name":319,"slug":320,"type":119},"Monitoring","monitoring",{"name":214,"slug":215,"type":119},"2026-08-20T03:53:13.51467",{"slug":324,"name":324,"fn":325,"description":326,"org":327,"tags":328,"stars":57,"repoUrl":56,"updatedAt":332},"braintrust-plan-agent-eval","plan evaluation workflows for LLM applications","Plan an evaluation workflow for an LLM application or agent by identifying the product decision, expected behavior, target population, evidence, datasets, scorers, validation, analysis, and release criteria. Use for broad or early-stage requests such as \"help me evaluate this agent,\" \"design an eval strategy,\" \"where do we start with evals,\" or when a user describes an agent, a product goal, or a production failure without naming an eval artifact. Do not use when the user already names a behavior spec, dataset, sample size, scorer, human review, experiment, analysis, release gate, or report — route to that skill.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[329,330,331],{"name":6,"slug":5,"type":119},{"name":27,"slug":26,"type":119},{"name":30,"slug":29,"type":119},"2026-08-20T03:52:55.941828",{"slug":334,"name":334,"fn":335,"description":336,"org":337,"tags":338,"stars":57,"repoUrl":56,"updatedAt":346},"braintrust-probe-capability-and-variability","probe system capability and output variability","Run the same dataset under controlled variants to measure either the ceiling of what a system can do or the spread of how reliably it does it. Use for questions about sensitivity to prompt paraphrases, formatting, ordering, seeds, or repeated runs; output consistency and agreement; whether a ranking survives a different prompt; flaky results; and equally for hidden, suppressed, sandbagged, or under-elicited capability, whether a low score means \"cannot\" or \"did not,\" or how prompting, demonstrations, scaffolding, tools, and fine-tuning compare. Do not use to find unknown failure modes or run adversarial attacks.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[339,342,343,344,345],{"name":340,"slug":341,"type":119},"Benchmarking","benchmarking",{"name":6,"slug":5,"type":119},{"name":27,"slug":26,"type":119},{"name":51,"slug":50,"type":119},{"name":30,"slug":29,"type":119},"2026-08-20T03:53:37.645766",{"slug":348,"name":348,"fn":349,"description":350,"org":351,"tags":352,"stars":57,"repoUrl":56,"updatedAt":358},"braintrust-red-team-agent","run adversarial red team testing","Plan and run adversarial worst-case testing of an authorized LLM or agent system against an explicit threat model, policy boundary, or high-impact prohibited behavior. Use for requests to red team, attack, jailbreak-test, or stress a system's guardrails, to assess prompt-injection or unsafe-tool-use exposure, or to turn adversarial findings into severity-ranked reports, mitigations, and regression cases. Do not use for open-ended discovery of ordinary failures, for measuring average quality, or without authorization over the target system.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[353,354,355],{"name":6,"slug":5,"type":119},{"name":27,"slug":26,"type":119},{"name":356,"slug":357,"type":119},"Security","security","2026-08-20T03:53:32.181496",{"slug":360,"name":360,"fn":361,"description":362,"org":363,"tags":364,"stars":57,"repoUrl":56,"updatedAt":372},"braintrust-report-eval-results","generate technical eval reports","Turn completed eval analysis into a technical report, release summary, chart set, or decision document with claims calibrated to the evidence. Use when writing up eval or benchmark results, preparing charts or tables of model comparisons, drafting a release note or blog post about eval numbers, or auditing a draft writeup for overstated claims, missing uncertainty, undisclosed search, or unpinned configuration. Do not use to compute the analysis itself or to decide whether to ship.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[365,366,367,368,369],{"name":36,"slug":35,"type":119},{"name":6,"slug":5,"type":119},{"name":45,"slug":44,"type":119},{"name":27,"slug":26,"type":119},{"name":370,"slug":371,"type":119},"Reporting","reporting","2026-08-20T03:53:33.675932",{"slug":374,"name":374,"fn":375,"description":376,"org":377,"tags":378,"stars":57,"repoUrl":56,"updatedAt":382},"braintrust-size-eval-dataset","calculate evaluation dataset sample sizes","Calculate or audit eval sample sizes, minimum detectable effects, confidence interval precision, required repeated runs, and clean-trial counts for bounding rare failures. Use when a user asks how many eval cases, items, scenarios, runs, or safety trials are needed, whether an existing dataset is adequately powered, whether N examples can detect an X-point gain, or how many clean trials certify a low violation rate. Account for paired designs, clustering, target confidence, and practical effect size. Do not use for general dataset composition.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[379,380,381],{"name":6,"slug":5,"type":119},{"name":27,"slug":26,"type":119},{"name":124,"slug":125,"type":119},"2026-08-20T03:53:32.932246",{"slug":384,"name":384,"fn":385,"description":386,"org":387,"tags":388,"stars":57,"repoUrl":56,"updatedAt":396},"braintrust-validate-eval-scorer","validate automated eval scorers","Validate automated eval scorers and LLM judges against expert-reviewed reference data. Use to compare scorer output with human labels, calculate agreement (kappa, alpha) with uncertainty, inspect confusion by class and severity, analyze subgroup failures, test shortcut and gaming cases, propagate scorer error into headline numbers, document blind spots, and decide whether a scorer is fit for exploration, trend monitoring, or release gating. Do not use to create the initial scorer or to design the human review workflow.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[389,390,391,392,393],{"name":6,"slug":5,"type":119},{"name":42,"slug":41,"type":119},{"name":27,"slug":26,"type":119},{"name":30,"slug":29,"type":119},{"name":394,"slug":395,"type":119},"Quality Assurance","quality-assurance","2026-08-20T03:52:59.694515",{"slug":398,"name":398,"fn":399,"description":400,"org":401,"tags":402,"stars":57,"repoUrl":56,"updatedAt":412},"braintrust-write-agent-behavior-spec","define and manage agent behavior specifications","Create, edit, audit, or compare versioned Agent Behavior specifications in .agents\u002Fbehaviors\u002F\u003Cname>\u002FBEHAVIOR.md. Use when a user wants to define recurring agent conduct, write a behavioral contract, make implicit trace-review expectations explicit, specify how an agent should handle uncertainty, destructive actions, or recovery, or describe intent, applicability, evidence, decision, execution, recovery, and failure modes. Also use to review or diff an existing BEHAVIOR.md. Do not use to implement system prompts, tool definitions, or scorer code.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[403,404,407,408,409],{"name":33,"slug":32,"type":119},{"name":405,"slug":406,"type":119},"Best Practices","best-practices",{"name":6,"slug":5,"type":119},{"name":45,"slug":44,"type":119},{"name":410,"slug":411,"type":119},"Technical Writing","technical-writing","2026-08-20T03:53:14.829935",{"slug":414,"name":414,"fn":415,"description":416,"org":417,"tags":418,"stars":57,"repoUrl":56,"updatedAt":426},"braintrust-write-eval-scorer","design and implement LLM eval scorers","Design, implement, edit, or audit narrow eval scorers for LLM applications and agents, including deterministic checks, reference or final-state comparisons, trace and tool-call checks, and anchored LLM-as-judge rubrics. Use when translating one observable criterion into scoring logic, choosing between deterministic and judge-based scoring, repairing a vague rubric, or defining handling for refusals, errors, timeouts, and parse failures. Do not use to validate scorer agreement against human labels or to design the human review workflow.",{"slug":5,"name":6,"logoUrl":7,"githubOrg":8},[419,420,421,424,425],{"name":33,"slug":32,"type":119},{"name":6,"slug":5,"type":119},{"name":422,"slug":423,"type":119},"Code Analysis","code-analysis",{"name":27,"slug":26,"type":119},{"name":30,"slug":29,"type":119},"2026-08-20T03:52:56.65537"]