index

Docs

Use Teammately docs as a layered manual: learning spine, operating tasks, reference library, recovery guides, and agent context.

Layered documentation

Start small, then drill into the exact operating layer.

The sidebar stays curated for first-time readers. Deep pages remain available through indexes, product-area pages, search, and agent-readable routes.

docs layers

learning manual
operating manual
reference library
recovery library
agent context layer

Task indexAll exact operations by workflow. Reference indexObjects, states, permissions, and source confidence. Troubleshooting indexRecovery by symptom and cause. Agent indexMachine-readable routes and safe agent boundaries.

Learning Manual Category education, product doctrine, lifecycle, boundaries, product loop, and first-start guidance. 7 pages Operating Manual Exact procedures for preparing cases, requesting contributions, governing standards, planning coverage, running evaluations, and preserving review context. 6 pages Coverage Engineering Public capability for designing what the benchmark must represent. 16 pages Correctness Elicitation Public capability for turning specialist judgment into executable standards. 1 pages Weave Public capability for constructing cases, variants, materials, and supported worlds. 1 pages Trialground Public capability for managed evaluation of harnesses and benchmark evidence. 1 pages Coevolve Public capability for evidence-backed candidate exploration and improvement. 1 pages Correctness Governance Project-level policies, applicability, rubrics, linked cases, provenance, and approval context. 3 pages Assets Reusable project cases, harnesses, project tools, worlds, weights, Comparison Directions, and Review Screens. 8 pages Agent Setup Project Context and Reference Materials. 3 pages Project Settings General, Regime, Project Input Schema, and Project Members. 5 pages Benchmark Datasets Benchmark cases, representation, and snapshots. 4 pages Coverage Management Benchmark setup, representation, Coverage Stories, Case Review, and Case Foundry. 5 pages Expert Contributions Benchmark-scoped expert requests, tasks, checkpoints, contributed artifacts, and status. 7 pages Benchmark Evaluations Version-scoped Runs, results, comparisons, Arena, rankings, and rubric evidence. 10 pages Improve Improvement Sessions, Goal Contracts, candidates, evaluation receipts, trajectories, and frontiers. 6 pages Governance and Reproducibility Human approval boundaries, AI feature boundaries, versioning, staleness, permissions, auditability, and reproducibility. 15 pages Admin Console Organization administration surfaces for members, groups, roles, security controls, usage, integrations, and API keys. 1 pages Public API Publication-gated customer contract for Case synchronization, external evaluation outputs, and benchmark result evidence. 0 pages Reference Library Object model, schemas, states, permissions, metadata, IDs, glossary, and source-confidence notes. 9 pages Recovery Library Symptom-driven recovery for uploads, outputs, cases, rubrics, policies, applicability, coverage, permissions, and benchmark runs. 19 pages Agent Context Layer Machine-readable surfaces, safe assumptions, forbidden assumptions, task maps, object maps, and retrieval boundaries. 1 pages