sincecutoff.com
/v1/json /v1/md llms.txt
← all changes

helmfile doctor analyzes diffs and gates high-risk changes

Introduced 2026-06-22 · score 0.95

HOW TO USE IT

Before applying planned changes, run `helmfile doctor` to review the diff; use `--output json` for a structured report. High-risk findings exit 2 unless `--force` is passed.

export HELMFILE_LLM_API_KEY="<api-key>" export HELMFILE_LLM_MODEL="<model-id>" helmfile doctor --output json
helmfile doctor --force --output json
helmfile doctor

WHY THIS HELPS

Review planned Helmfile changes in a CI pipeline before applying them.

REQUIREMENTS AND LIMITATIONS

  • Version: v1.6.0
  • --output <format>: Doctor report format; `json` emits structured JSON. Default is text.
  • --force: Skip the exit-2 gate when a high-severity risk is found.
  • --context <N>: Diff context lines; doctor defaults to 3.
  • llm.baseURL: OpenAI-compatible endpoint base URL; defaults to `https://api.openai.com/v1`.
  • llm.apiKey (required): API key; required to enable LLM analysis.
  • llm.model (required): Model identifier; required to enable LLM analysis.
  • llm.timeout: Per-request timeout; default 60s.
  • llm.maxTokens: Maximum completion tokens; default 4096.
  • --llm-base-url <URL>: Set the OpenAI-compatible endpoint URL.
  • --llm-api-key <key>: Set the LLM API key.
  • --llm-model <id>: Set the model identifier.
  • --llm-timeout <duration>: Set per-request timeout; Go duration format.
  • --llm-max-tokens <N>: Set completion token cap.
  • --diff-output <format>: Set helm-diff plugin output format; `--output` is reserved for doctor report format.
  • --show-secrets: Ignored by doctor; secrets remain redacted.
  • --suppress-secrets: Drop Secret resource structure from the diff.
  • Prerequisite: LLM analysis requires `apiKey` and `model`; `baseURL` is optional. The endpoint must speak OpenAI Chat Completions protocol.
  • Prerequisite: `helmfile diff` requires helm-diff plugin v2.9.0+1 or greater.
  • The endpoint must be OpenAI-compatible and support Chat Completions.
  • The default `--context` value is 3.
  • A high risk exits with code 2 unless `--force` is passed; other errors exit with code 1.
  • An LLM call failure degrades to exit code 0.
  • Secrets are always redacted before any byte leaves the process; `--show-secrets` is silently ignored.
  • Without LLM configuration, doctor degrades to a plain `helmfile diff` with `--show-secrets` forced off.
  • Most `helmfile diff` flags are accepted for parity.

FIRST-PARTY EVIDENCE

Check the installed version and the source before using it. Entry JSON

Report a problem with this card