ContractSpec docs

OSS-first docs

These docs teach the open system first: contracts, generated surfaces, runtimes, governance, and incremental adoption. Studio shows up as the operating layer on top, not as the source of truth.

AI index

Eval harness — prompt fixtures + assertions

M5 P4 eval harness: 5 domain fixtures × 4 graphOps tools with tolerance-aware mutation assertions and per-tool latency/cost budgets.

field.key.label
Eval harness — prompt fixtures + assertions
field.version.label
field.type.label
field.title.label
Eval harness — prompt fixtures + assertions
field.description.label

M5 P4 eval harness: 5 domain fixtures × 4 graphOps tools with tolerance-aware mutation assertions and per-tool latency/cost budgets.

field.tags.label
agent,eval,graphOps,llm,m5
field.owners.label
field.stability.label
internal

Defines types for the M5 P4 eval gate across 5 template domains: workflow-design | agent-tree | conversation-routing | engineering-loop | person-activity

Each EvalFixture bundles:

•

EvalPromptFixture: id, domain, tool, prompt, contextGraphId, description

•

EvalMutationAssertion[]: tolerance-gated checks on tool output

•

EvalBudget: maxLatencyMs + GRAPH_AI_BUDGETS values for the tool

Tolerance in [0, 1]: 0 = exact match required, 1 = any output accepted.