`maida baseline`#

Captures a baseline snapshot from a completed run. The snapshot records structural metrics (event counts, tool path, token usage, duration, etc.) that maida assert can later compare against. See Regression testing for the full workflow.

Usage:

maida baseline [TRACE_ID] [--out PATH] [--force]
maida baseline --from-report REPORT.json [--out PATH] [--force]

Arguments / options:

Argument/Option	Default	Description
`TRACE_ID`	(latest run)	OTel trace ID or prefix for legacy single-run capture
`--from-report`	-	Build an immutable multi-trial sample from report v2; mutually exclusive with `TRACE_ID`
`--out`, `-o`	`.maida/baselines/<run_name>.json`	Output path for the baseline JSON file
`--force`	`false`	Explicitly allow replacing an existing baseline

Examples:

maida run my_agent.py --trials 25 --no-fail-fast --json-out report.json
maida baseline --from-report report.json --out .maida/baselines/my_agent.json
maida baseline a1b2c3d4 --out baselines/legacy-single-run.json

Existing files are preserved by default. To deliberately replace one, repeat the command with --force. Use maida accept to accept a run change with a recorded reason and prior baseline hash.

Exit codes: 0 success; 2 run not found, invalid input, or output already exists; 10 internal error.

Report-based capture stores the raw per-trial numeric and invariant vectors, an environment fingerprint, and structural signatures deduplicated with counts. The sample is immutable and never accumulates across gate runs. Check it into version control as part of the reviewed diff.

maida baseline#

`maida baseline`#