Skip to main content
A KHAW agent report is useful only if another agent or human can reproduce it.

​
Minimum proof packet

Command: <exact command or URL>
Exit status: <0/non-zero or HTTP status>
Output: <sanitized relevant output>
Proves: <one sentence>
Does not prove: <important limitation>
Next safe action: <command/request>
Mutation: none | local-only | approved shared target

​
Good proof

Command: khal install . --target dev --dry-run --json
Exit status: 0
Output: summary.status=pass; mutation=none
Proves: app manifest/source can pass dev install preflight.
Does not prove: the app is installed or healthy in dev.
Next safe action: ask for GO to run non-dry-run install or keep local.

​
Bad proof

It should work now.

​
Evidence destinations

  • Short task: paste back to the chat.
  • Multi-step work: /wish new <slug> and keep artifacts under Brain.
  • Architecture decision: /brainstorm and preserve council output.
  • UI proof: screenshot or browser output plus console cleanliness.
  • Release proof: PR/commit, CI, deploy ledger, version/image/source evidence.
TASK: I am reading `khaw/agent-evidence.mdx` (Agent evidence and proof). Use this page as the contract, then verify current CLI/output before you guide me.
CONTEXT: I may be a new KHAL FDE. Prefer read-only checks and dry-runs first. Do not mutate customer, HML, production, credentials, SSH, Gitea, or model-provider state without an explicit GO.
SAFE FIRST COMMANDS: Check versions, identity, target, git source/ref, KHAW doctor/status, KHAL context, and dry-run output. Redact secrets and private URLs.
EVIDENCE: Return command, exit status, sanitized output, what it proves, and the next safe action.