cliare-cli-agent-readiness

Run cliare measure against the CLI binary with a profile such as standard or deep, then use cliare summary to view the readiness scorecard. The measurement writes a command index, issue ledger, and scorecard to the output directory.

Use cliare report security after a measurement to review undocumented filesystem writes, network calls, and credential access. The report highlights discovery commands that appear safe but quietly modify state.

CLIARE measures released CLI binaries as black boxes, so it works with any executable regardless of implementation language. You can adjust depth, probe budget, and concurrency for CLIs with very large command surfaces.

Yes, install cliare via the install script in a workflow step, run cliare measure against your built binary, and upload the generated artifacts. It also produces findings.sarif for GitHub code scanning and junit.xml for test reporting.

Missing commands usually mean the probe depth or budget was too low. Switch to the deep profile or increase --max-depth and --max-probes, and use --refresh to force regeneration of artifacts.

Black-box probing cannot see internal code paths, so commands requiring auth, network, fixtures, or daemons may show lower confidence unless measured with the appropriate context flags. Very large surfaces also require explicit probe budgets to complete in reasonable time.