---
source: crates/perfgate-cli/tests/cli_help_snapshot_tests.rs
expression: "help_output(&[\"--help\"])"
---
Perf budgets and baseline diffs for CI / PR bots
Usage: perfgate [OPTIONS] <COMMAND>
Commands:
run Run a command repeatedly and emit a run receipt (JSON)
compare Compare a current receipt against a baseline and emit a compare receipt (JSON)
md Render a Markdown summary from a compare receipt
github-annotations Emit GitHub Actions annotations from a compare receipt
export Export a run or compare receipt to CSV, JSONL, HTML, Prometheus, or JUnit format
promote Promote a run receipt to become the new baseline
ratchet Preview or apply conservative budget ratcheting from compare evidence
report Generate a cockpit-compatible report from a compare receipt
check Config-driven one-command workflow
doctor Diagnose local setup, config, baselines, artifacts, CI, and server reachability
paired Run paired benchmark: interleave baseline and current commands for reduced noise
baseline Inspect local baselines and manage baselines on the baseline server
admin Administer baseline service operations
audit List and export baseline service audit events
summary Summarize one or more compare receipts in a terminal table
aggregate Aggregate multiple run receipts (e.g. from a fleet) into a formal aggregate receipt
decision Evaluate scenario and tradeoff evidence into a review-ready decision summary
probe Compare named probe receipts and emit probe-level deltas
scenario Evaluate configured workload scenarios from compare receipts
tradeoff Evaluate configured tradeoff rules against scenario evidence
bisect Automatically find the commit that introduced a performance regression
cargo-bench Wrap `cargo bench` and produce perfgate run receipts
blame Analyze changes in Cargo.lock to identify dependency updates causing binary size regressions
fleet Fleet-wide dependency regression analysis across projects
explain Provide AI-ready prompts and automated playbooks for diagnosing performance regressions
ingest Import benchmark results from external frameworks into perfgate format
badge Generate an embeddable SVG status badge from a report or compare receipt
discover Auto-detect benchmarks in a repository without manual configuration
diff Quick "did I make it slower?" comparison against baselines
init Scan a repository and generate a perfgate.toml config file
watch Watch for file changes and re-run benchmarks with live terminal output
serve Start a local dashboard server backed by SQLite
scale Validate computational complexity (scaling behavior) of a benchmark
comment Post or update a performance report comment on a GitHub pull request
trend Analyze metric trends and predict budget threshold breaches
help Print this message or the help of the given subcommand(s)
Options:
-h, --help Print help
-V, --version Print version
Global Options:
--baseline-server <BASELINE_SERVER> URL of the baseline server (e.g., http://localhost:3000/api/v1) Can also be set via PERFGATE_SERVER_URL environment variable
--api-key <API_KEY> API key for authentication with the baseline server. Can also be set via PERFGATE_API_KEY environment variable
--project <PROJECT> Project name for multi-tenancy. Can also be set via PERFGATE_PROJECT environment variable