Appearance

skipprd doctor ​

skipprd doctor checks your configuration before you run anything long. It loads skippr.yml, resolves every ${NAME} reference, and checks that each pipeline has a source, that every reference between pipelines and entries resolves, and that paired sinks match. It does not connect to your source or destination, so it is fast and safe to run anywhere: after editing the config, in CI on every change, or as the first step of a deploy. Doctor checks every pipeline in the file.

Usage ​

python
import skippr

session = skippr.Session(skippr.Config.discover().get_pipeline("bikehire"))
result = session.doctor()
print(result["ok"])
bash
skipprd doctor [--output text|json] [--config <path>]

Options ​

FlagDefaultDescription
--output <MODE>texttext prints one line per check. json prints {"ok": bool, "checks": [...]} for scripts.
--config <PATH>./skippr.ymlConfig file to check.
--log [LEVEL]OffPrint logs to stderr.

See CLI overview for the other global flags.

What doctor checks ​

CheckPasses when
Config loadsskippr.yml parses and every ${NAME} reference has a value. If not, doctor stops with [skippr] config failed: ... before running other checks.
PipelinesAt least one pipeline is configured.
Sourcesdata_sources is not empty.
DestinationsAlways passes. A pipeline without a data_sink is valid: its data stays in the WAL.
Each pipelineIts data_source, data_sink, and other references point at entries that exist.
Whole configNo two SkipprLake or DuckDB destinations write the same namespace, and paired sinks (SkipprLake, AthenaIceberg, DuckDB) match their schema sink.
Storage modeReports skipprd_el_storage_mode for each pipeline.

Doctor validates configuration only. To test credentials and connectivity, run skipprd discover, which reads the source but never writes to the destination.

Examples ​

Check the config in the current directory ​

bash
skipprd doctor
text
  [ok]   1 pipeline(s) configured
  [ok]   data_sources configured
  [ok]   data_sinks configured (optional)
  [ok]   pipeline 'bikehire' (source+sink)
  [ok]   skipprd_el_storage_mode=local

All checks passed.

A failing check is marked [FAIL] and the summary reads Some checks failed.

JSON for CI ​

bash
skipprd doctor --output json
json
{
  "ok": true,
  "checks": [
    { "ok": true, "severity": "info", "message": "1 pipeline(s) configured" },
    { "ok": true, "severity": "info", "message": "data_sources configured" },
    { "ok": true, "severity": "info", "message": "no data_sink; WAL is the dataset" },
    { "ok": true, "severity": "info", "message": "pipeline 'rides' (source, WAL-only)" },
    { "ok": true, "severity": "info", "message": "skipprd_el_storage_mode=local" }
  ]
}

Each check has ok, severity (info or error), and message. A check may also include suggested_fix_command.

Fail a CI job on a bad config ​

bash
skipprd --config deploy/skippr.yml doctor --output json > doctor.json || {
  jq -r '.checks[] | select(.ok == false) | .message' doctor.json
  exit 1
}

Doctor needs every ${NAME} variable to be set, so export your secrets (or test placeholders) in the CI job first.

Troubleshooting ​

OutputCauseFix
[skippr] config failed: ... references ${NAME} at ... but that environment variable is not setA secret reference has no value.export NAME=..., or add it to a .env file next to skippr.yml.
[FAIL] no pipelines configured and [FAIL] data_sources is emptyDoctor found no config file, or the file is empty.Run from the folder that contains skippr.yml, or pass --config.
[FAIL] ... must be equal ...A paired data sink and its schema sink differ.Make the two blocks identical, or re-add them with skipprd connect.
[skippr] config failed: Invalid Skippr configuration in ...skippr.yml has a YAML error or an unknown key.Fix the line named in the message.

Exit codes ​

CodeMeaning
0All checks passed.
1At least one check failed, or the config could not be loaded.

Next steps ​