Auggie CLI vs Harness²

Auggie CLI

Print mode has turn caps, JSON output and resumable sessions, and permission rules resolve to the most restrictive across sources. Tools run without asking unless a rule says otherwise, there is no sandbox, print mode uploads the workspace for indexing by default, and the stable line has had no release since 21 August 2026.

Harness²

Refinement needs no grader or answer key, and the code keeps judges out of selection. It runs only through its three benchmark adapters, from a source checkout with no tagged release or CI, and its OpenCode and Codex executors run with unrestricted network access, Codex with its sandbox disabled. The benchmark gains are the authors' claims.

Auggie CLI

1. Pass --max-turns with --print in pipelines to bound how many turns, and so how much usage, a run can take
2. Commit deny rules to .augment/settings.json before an unattended run. Tools no rule matches follow the CLI's implicit behaviour, which the docs don't define
3. Add secrets and data directories to .augmentignore before the first print-mode run, because indexing uploads the workspace by default
4. Use a service account token in CI on team and Enterprise plans. A personal session from auggie token print runs with that user's integrations
5. Pin the version in CI. latest is 0.x, while preview and daemon carry 1.x prereleases whose permission rules ignore some legacy tool names

Harness²

1. Don't plan to call Harness² as a service. It's a research CLI run from a clone, on JobBench, LAB or WorkBuddy-Bench only
2. Ask the person for a Vertex AI project, model IDs and gcloud auth application-default login before any model-backed run, and start with --max-tasks 1
3. Run executors behind an egress firewall and block huggingface.co and github.com after setup, as the docs advise, so rubrics can't be fetched
4. Write .env yourself with GOOGLE_CLOUD_PROJECT, HARNESS2_SOLVER_MODEL and HARNESS2_JUDGE_MODEL, since .env.example isn't shipped, and never commit it
5. Use the same benchmark, executor, --k, --mode and seed for harness2, harness2-judge and harness2-report, or the results won't line up