Experiments

AgentV eval files are the only runnable authoring artifact. The eval file defines the experiment. Use top-level name as the result namespace, top-level target for the system under test, and top-level policy for runtime and gating controls such as run count, timeout, budgets, and thresholds. AgentV does not have a separate experiment.yaml file, top-level run_group, or schema-significant experiments/ directory.

name: support-regression
target: codex-gpt5
model: gpt-5-codex
policy:
  runs: 4
  timeout_seconds: 720
  budget_usd: 2.00

workspace:
  hooks:
    before_all:
      command: ["bash", "-lc", "bun install && bun run build"]

tests:
  - id: refund-eligibility
    input: Can this customer get a refund?
    criteria: Applies the refund policy correctly

experiment: was a prerelease authoring block and is rejected. Move experiment.target to top-level target, experiment.model to top-level model, and runtime controls to top-level policy with runs, timeout_seconds, threshold, and budget_usd. execution: is accepted only as a legacy top-level runtime alias for existing eval files and target matrices.

Layout Conventions

Use directories for human organization, not schema behavior. A common layout is:

evals/
  suites/
    refunds.eval.yaml
  cases/
    refund-smoke.cases.yaml
experiments/
  refunds-codex.eval.yaml

In that layout, evals/suites/refunds.eval.yaml is a reusable task suite, evals/cases/refund-smoke.cases.yaml is raw case data, and experiments/refunds-codex.eval.yaml is a wrapper eval. The wrapper still runs only because it is eval YAML:

experiments/refunds-codex.eval.yaml

name: refunds-codex
target: codex-gpt5

tests:
  - id: local-edge-case
    input: Check a damaged final-sale refund.

imports:
  suites:
    - path: ../evals/suites/refunds.eval.yaml
  tests:
    - path: ../evals/cases/refund-smoke.cases.yaml

The experiments/ folder is optional and user-owned. AgentV does not scan it for special files or infer runtime behavior from the path; the same wrapper eval could live under evals/wrappers/, benchmarks/, or beside the suite it runs.

Suite And Test Imports

Use imports.suites for full child suites and imports.tests for raw test rows. Inline tests remain raw cases owned by the current file.

imports:
  suites:
    - path: evals/support/*.eval.yaml
      select:
        test_ids:
          - refund-*
          - missing-order-date
        tags: regression
        metadata:
          priority: high
      run:
        threshold: 1.0
        timeout_seconds: 300
  tests:
    - path: cases/*.cases.yaml
    - path: cases/regression.jsonl

tests:
  - cases/smoke/*.cases.yaml

imports.suites preserves the imported suite’s task contract: metadata, workspace, shared input, shared assertions, and tests. The parent eval still owns the single run bundle and runtime policy. Child suite legacy runtime blocks are ignored when imported; use parent target/policy for run policy and import run: for scoped threshold, timeout, or budget overrides.

A parent eval that imports any imports.suites entry must not define top-level workspace. Imported suites own task environment. If the parent should provide workspace context, import raw cases with imports.tests or shorthand paths instead of importing an eval suite.

imports.tests imports only raw test entries. It intentionally drops shared context from an imported eval suite, so parent suite fields apply to those raw cases.

Import select.test_ids filters imported test IDs with glob patterns. Import select.tags filters each imported case’s effective metadata.tags. Effective case tags are suite-first and deduped: suite.tags + suite.metadata.tags + test.metadata.tags. Top-level suite tags still remain suite identity metadata for discovery and reporting; selection reads the merged case metadata view. Import select.metadata filters case metadata by key/value, where selector values may be scalars or lists. Globbed include paths are resolved in deterministic path order, then test order.

String-valued tests and string entries inside tests[] are raw-case import shorthand. They are equivalent to imports.tests and may point at raw case files, directories, or globs. Importing another eval suite must use imports.suites.

Suite imports are resolved as a deterministic include graph. Circular imports.suites imports fail validation with the import chain; raw-case shorthand does not recursively load suite runtime blocks.

Imported suite rows keep their source suite metadata in index.jsonl. Use each row’s result_dir as the authoritative path to generated artifacts inside the run directory; do not infer layout from suite names.

Scoped Run Overrides

Use scoped run: blocks for result interpretation and scheduling policies that vary by include group or test case. Precedence is:

test.run > import run > parent policy

target: agent
policy:
  threshold: 0.8
  runs: 3

imports:
  suites:
    - path: ./evals/flaky-agentic/**/*.eval.yaml
      select:
        tags: [agentic]
      run:
        timeout_seconds: 300

    - path: ./evals/regression/**/*.eval.yaml
      select:
        tags: [must-pass]
      run:
        threshold: 1.0
        timeout_seconds: 300

tests:
  - id: critical-case
    input: "..."
    criteria: Must pass exactly
    run:
      threshold: 1.0
      budget_usd: 0.50

Scoped run: supports threshold, timeout_seconds, and budget_usd for public eval authoring. Candidate-changing fields such as target and targets stay parent-level. Workspace mutation belongs in workspace.hooks, and runner-specific setup belongs in targets[].hooks.

Lifecycle Ownership

policy: configures evaluation policy. It does not own commands that prepare files, dependencies, repos, or target-specific runner state.

Need	Put it in
Install dependencies, build the repo, seed files	workspace.hooks.before_all
Reset or apply per-case state	workspace.hooks.before_each / workspace.hooks.after_each
Configure an agent runner or provider variant	targets[].hooks
Choose the target	top-level target
Override the target’s default model	top-level model
Configure run count, budget, timeout, threshold	top-level policy
Bind an existing local workspace directory	--workspace-path or .agentv/config.local.yaml

workspace:
  hooks:
    before_all:
      command: ["bash", "-lc", "bun install && bun run build"]

targets:
  - name: agent-with-skills
    provider: codex
    hooks:
      before_each:
        command: ["sh", "-c", "cp -R skills \"{{workspace_path}}/.codex/skills\""]

target: agent-with-skills
policy:
  runs: 3

experiment.workspace and top-level experiment: are not authored eval YAML fields. Existing local workspace paths are machine-local bindings: pass --workspace-path for a one-off run or put execution.workspace_path in .agentv/config.local.yaml. Put repos, templates, hooks, Docker config, env checks, and isolation under top-level or case-level workspace.

Repeat Runs

Use policy.runs when you want AgentV to try each case more than once:

policy:
  runs: 3

The current repeated-attempt aggregation treats the case as successful when any attempt succeeds. AgentV does not expose this as pass@k; in ML and code-generation evaluation, pass@k is a statistical metric with different semantics.

Result Layout

Eval runs write to the selected result bucket:

.agentv/results/<experiment>/<timestamp>/

CLI --experiment sets the bucket explicitly. Without that flag, AgentV uses top-level name from the eval file, then the eval filename. The Dashboard still uses “Experiment” for the comparison and result grouping concept; folder names are only storage allocation and must not define result semantics.

Imported source suite metadata appears in index.jsonl rows and manifests. Use index.jsonl fields such as eval_path, test_id, target, and result_dir for identity and artifact discovery instead of reconstructing paths from suite names or wrapper layout.

For the complete result file contract, including why row metadata is semantic truth and directories are storage allocation, see Result Artifact Contract.