{"query":"Improve","corpusVersion":"local","generatedAt":"2026-09-13T04:31:55.636Z","results":[{"blockId":"improve.overview#improve","pageId":"improve.overview","title":"Improve","pageTitle":"Improve","url":"https://teammately.ai/docs/improve.md","humanUrl":"https://teammately.ai/docs/improve#improve","markdownUrl":"https://teammately.ai/docs/improve.md","sectionId":"improve","kind":"concept","productArea":"improve","score":424.60116982220865,"reasons":["search_match","title_match","display_title_match","summary_match","term_match","prefix_or_fuzzy_match"],"markdown":"# Improve\n\nImprove is the benchmark-scoped workspace for coordinating candidate work from benchmark evidence. Each Improvement Session pins the exact starting evidence it uses, including any Benchmark Version, Harness Version, and Run identities. Sessions preserve the goal, authorization, agent and worker activity, candidate Harness Versions, canonical evaluation receipts, and Current frontier."},{"blockId":"improve.start-session#start-an-improvement-session","pageId":"improve.start-session","title":"Start an Improvement Session","pageTitle":"Start an Improvement Session","url":"https://teammately.ai/docs/improve/start-improvement-session.md","humanUrl":"https://teammately.ai/docs/improve/start-improvement-session#start-an-improvement-session","markdownUrl":"https://teammately.ai/docs/improve/start-improvement-session.md","sectionId":"start-an-improvement-session","kind":"task","productArea":"improve","score":218.09044680444958,"reasons":["search_match","title_match","display_title_match","term_match","prefix_or_fuzzy_match"],"markdown":"# Start an Improvement Session\n\nStart an Improvement Session when evaluation evidence justifies a candidate change or bounded investigation. The setup should turn a free-form intention into a measurable Goal Contract before work begins.\n\nChoose the saved Harness Version and a specific baseline launch. A baseline may contain one Run. Set the number of Runs for future candidate evaluations independently; a difference from the baseline count is informational. The work forecast and authorized budget use the chosen candidate count."},{"blockId":"improve.overview#related-workflows","pageId":"improve.overview","title":"Related workflows","pageTitle":"Improve","url":"https://teammately.ai/docs/improve.md","humanUrl":"https://teammately.ai/docs/improve#related-workflows","markdownUrl":"https://teammately.ai/docs/improve.md","sectionId":"related-workflows","kind":"concept","productArea":"improve","score":154.6087842913911,"reasons":["search_match","page_title_match","summary_match","term_match","prefix_or_fuzzy_match"],"markdown":"## Related workflows\n\n{% related-card-grid title=\"Related workflows\" %}\n- [Start an Improvement Session](/docs/improve/start-improvement-session)\n- [Write and confirm a Goal Contract](/docs/improve/goal-contracts)\n- [Choose Work or Evolve](/docs/improve/work-and-evolve)\n- [Interpret candidates and the current frontier](/docs/improve/candidates-and-frontier)\n- [Read chronology and trajectories](/docs/improve/chronology-and-trajectories)\n- [Inspect evaluation results](/docs/benchmark-evaluations/inspect-results)\n- [Request an Expert Contribution](/docs/expert-contributions/request-contribution)\n{% /related-card-grid %}"},{"blockId":"improve.overview#source-confidence","pageId":"improve.overview","title":"Source confidence","pageTitle":"Improve","url":"https://teammately.ai/docs/improve.md","humanUrl":"https://teammately.ai/docs/improve#source-confidence","markdownUrl":"https://teammately.ai/docs/improve.md","sectionId":"source-confidence","kind":"concept","productArea":"improve","score":148.3851853675992,"reasons":["search_match","page_title_match","summary_match","term_match","prefix_or_fuzzy_match"],"markdown":"## Source confidence\n\nCode-backed: the active versioned Improve workspace and session contract define starting evidence, Goal Contracts, Work and Evolve behavior, candidates, receipts, chronology, frontiers, and lifecycle controls. Internal coordinator implementation is supporting evidence, not a public API."},{"blockId":"improve.overview#candidates-and-current-frontier","pageId":"improve.overview","title":"Candidates and current frontier","pageTitle":"Improve","url":"https://teammately.ai/docs/improve.md","humanUrl":"https://teammately.ai/docs/improve#candidates-and-current-frontier","markdownUrl":"https://teammately.ai/docs/improve.md","sectionId":"candidates-and-current-frontier","kind":"concept","productArea":"improve","score":147.79646311463765,"reasons":["search_match","page_title_match","summary_match","term_match","prefix_or_fuzzy_match"],"markdown":"## Candidates and current frontier\n\nA candidate should identify its Harness version, hypothesis, parent or branch context, and evaluation evidence. The current frontier represents retained candidates supported by the latest valid evidence and Goal Contract. It is not synonymous with the most recent candidate or the highest aggregate score.\n\nEvaluation receipts link candidates back to Benchmark Evaluations. If a candidate improves one objective but violates a constraint or regresses a must-level rubric, the frontier should keep that tradeoff visible."},{"blockId":"improve.overview#definition","pageId":"improve.overview","title":"Definition","pageTitle":"Improve","url":"https://teammately.ai/docs/improve.md","humanUrl":"https://teammately.ai/docs/improve#definition","markdownUrl":"https://teammately.ai/docs/improve.md","sectionId":"definition","kind":"concept","productArea":"improve","score":147.22673433989632,"reasons":["search_match","page_title_match","summary_match","term_match","prefix_or_fuzzy_match"],"markdown":"## Definition\n\nA new session begins with **Start from evidence**. The operator selects the target and relevant benchmark evidence, then prepares a **Goal Contract**. The contract binds the free-form objective to canonical target identities, measurement bindings, constraints, unresolved items, and intervention boundaries.\n\nThe current Improve workspace offers **Coevolve**, **External Agents**, and **External fine-tuning** experiences where enabled. Coevolve supports bounded Work and Evolve behavior; external experiences use explicit handoff or provider boundaries. All retained candidates still rely on observable Harness Versions and canonical Benchmark Evaluations rather than unrecorded claims of progress."},{"blockId":"improve.overview#decision-checkpoint","pageId":"improve.overview","title":"Decision checkpoint","pageTitle":"Improve","url":"https://teammately.ai/docs/improve.md","humanUrl":"https://teammately.ai/docs/improve#decision-checkpoint","markdownUrl":"https://teammately.ai/docs/improve.md","sectionId":"decision-checkpoint","kind":"concept","productArea":"improve","score":142.11401831575984,"reasons":["search_match","page_title_match","summary_match","term_match","prefix_or_fuzzy_match"],"markdown":"## Decision checkpoint\n\n| Need | Session choice | Required evidence |\n| --- | --- | --- |\n| Investigate or implement one bounded direction | Work | Confirmed Goal Contract and pinned failures or target evidence |\n| Explore several competing hypotheses | Evolve | Comparable measurement plan and explicit exploration authorization |\n| Delegate implementation externally | Worker package | Pinned target, return contract, and observable returned candidate |\n| Decide which candidates remain viable | Current frontier | Canonical evaluation receipts and goal constraints |\n| Stop scheduling new work temporarily | Pause | Retained evidence and chronology remain available |\n| Correct missing benchmark meaning | Contribution or coverage follow-up | Upstream policy, rubric, case, or coverage observation is named |"},{"blockId":"improve.overview#goal-contract-and-chronology","pageId":"improve.overview","title":"Goal Contract and chronology","pageTitle":"Improve","url":"https://teammately.ai/docs/improve.md","humanUrl":"https://teammately.ai/docs/improve#goal-contract-and-chronology","markdownUrl":"https://teammately.ai/docs/improve.md","sectionId":"goal-contract-and-chronology","kind":"concept","productArea":"improve","score":142.11401831575984,"reasons":["search_match","page_title_match","summary_match","term_match","prefix_or_fuzzy_match"],"markdown":"## Goal Contract and chronology\n\nThe Goal Contract is the session's authority boundary. Confirm it only after the target and measurement plan are resolvable. Later revisions should remain explicit rather than silently changing what earlier candidates attempted.\n\nChronology records durable session transitions, agent activity, proposed approaches, candidates, evaluations, pauses, resumptions, cancellation, and terminal state. Narrated trajectories expose bounded work segments and can identify unannounced durable activity without pretending to know private external-worker behavior."}]}