Belvedir CLI

npm install -g @belvedir/cli
belvedir login          # opens your browser (Google sign-in)
belvedir projects

belvedir whoami                      # the signed-in account
belvedir orgs                        # organizations, roles, balances
belvedir projects                    # projects you can access
belvedir keys my-agent               # your API keys in a project
belvedir conversations my-agent --n 50
belvedir traces my-agent             # raw spans (name, model, status)
belvedir tasks my-agent
belvedir groups my-agent             # training sets
belvedir cleaning my-agent           # the cleaning log
belvedir models                      # fine-tuned models
belvedir runs my-agent               # training + optimization runs
belvedir benchmarks
belvedir benchmarks show <run-id>    # status, score, log tail
belvedir environments my-agent       # generated environments + simulations
belvedir review my-agent             # proposals awaiting review
belvedir setup my-agent              # target / loops / design / autonomy
belvedir routing my-agent
belvedir members my-org              # roster + pending invites
belvedir access my-agent             # per-member project access
belvedir balance
belvedir billing my-org              # the org's credit ledger
belvedir github                      # connected GitHub accounts

belvedir projects create my-agent
belvedir projects rename my-agent "my agent v2"
belvedir projects delete my-agent --yes
belvedir keys revoke <key-id>
belvedir keys rename <key-id> production
belvedir train my-agent --method sft      # launch a LoRA training run (sft | opsd | rl)
belvedir train my-agent --method rl --reward similarity --algorithm auto
belvedir train cancel <run-id>
belvedir analyze my-agent                 # queue an optimization run
belvedir review approve <analysis-id>     # opens the PR (or reject)
belvedir setup my-agent --target "Fewer failed orders" --loops harness,lora --codebase https://github.com/you/agent --autonomy harness=review --design loop2 --keys all
belvedir autotrain my-agent --traces 500  # or --days 7, or --off
belvedir benchmarks create my-agent --name smoke --git-url https://github.com/you/harness --run "python run.py"
belvedir benchmarks add my-agent swe-bench-verified --env MODEL=anthropic/claude-sonnet-5,MODAL_TOKEN_ID=...,MODAL_TOKEN_SECRET=...
belvedir benchmarks catalog               # the public presets and their env
belvedir benchmarks run <harness-id> --attempts 5 --grading pass@k
belvedir benchmarks run <harness-id> --model <trained-model-id>   # GPU eval of a fine-tune
belvedir benchmarks cancel <run-id>
belvedir benchmarks delete <harness-id> --yes
belvedir environments generate my-agent --group <group-id> --preset belvedir   # the default first-party driver
belvedir environments export my-agent --id <environment-id> --out ./env       # Harbor dataset: harbor run -p ./env
belvedir environments generate my-agent --group <group-id> --preset belvedir --synthesize   # + gated rubric verifiers
# --synthesize starts the pass on the worker and polls it (6–10 minutes; --no-wait returns at once).
# Needs @belvedir/cli 0.4.0 or newer; older versions hold one request open and print an error partway
# through while the environment still lands.
belvedir environments status <environment-id> [--wait]   # a synthesis's progress or gate report
belvedir environments generate my-agent --group <group-id> --repo https://github.com/you/agent --run "python eval.py"
belvedir environments simulate my-agent --group <group-id>
belvedir environments declare my-agent --name "Refund requests" --description "Customer asks for a refund..."
belvedir signals my-agent                 # error/slow/frustration scan
belvedir validation my-agent              # success rates + shipped deltas
belvedir routing stats my-agent           # per-group model records + recommendation
belvedir simulations run <simulation-id> --episodes 8
belvedir harness map my-agent you/agent   # map a connected repo's harness
belvedir repos add my-agent you/agent     # project repository access
belvedir models download <model-id>       # signed adapter URL
belvedir deploy <model-id>                # Together deploy, start to endpoint
belvedir billing payg my-org              # save the org's card (auto reload / pay as you go)
belvedir routing set my-agent --default qwen3:4b --strong zai-org/GLM-5.2-FP8 --small Qwen/Qwen3.5-9B --threshold 0.5
belvedir routing pin my-agent <group-id> qwen3:4b
belvedir routing unpin my-agent <group-id>
belvedir routing routers my-agent         # all routers: auto + drafts
belvedir routing promote <router-id>      # switch which router serves traffic
belvedir orgs create "my team"
belvedir members invite my-org teammate@example.com
belvedir access set my-agent <user-id> off
belvedir outcome my-agent <session-id> fail
belvedir billing buy my-org --amount 20   # opens Stripe Checkout
belvedir billing redeem my-org WELCOME
belvedir github connect                   # opens the GitHub App install page