Update gunshi to 0.35; nested cache subcommands, positional & hidden args#799
Open
theoephraim wants to merge 6 commits into
Open
Update gunshi to 0.35; nested cache subcommands, positional & hidden args#799theoephraim wants to merge 6 commits into
theoephraim wants to merge 6 commits into
Conversation
- bump gunshi + plugin-completion to 0.35.1, pin plugin-i18n peer - make `varlock cache status`/`clear` real nested subcommands - declare positional args on printenv/explain/reveal/scan/audit - hide deprecated `run --no-inject-graph` from help (still works) - docs: document `run --inject` instead of the deprecated flag - smoke tests for the new positionals + --no-inject-graph back-compat - smoke-tests: drop dead captureOutput option, fix stale README
Contributor
|
The changes in this PR will be included in the next version bump.
|
Deploying with
|
| Status | Name | Latest Commit | Preview URL | Updated (UTC) |
|---|---|---|---|---|
| ✅ Deployment successful! View logs |
varlock-website | ad96ba0 | Commit Preview URL Branch Preview URL |
Jun 18 2026, 04:57 AM |
commit: |
- document previously-missing `explain` and `install-plugin` commands - document exit codes for load/run/printenv/scan (audit already had them) - document the load JSON output contract (json / --agent / json-full) and the stdout/stderr summary flags for machine consumers - cross-link machine-readable output + exit codes in the AI Tools guide - encrypt: add EXAMPLES (incl. stdin usage), fix doubled default in --key-id help, document --key-id in the reference
The varlock("local:...") reference doesn't encode a keyId and the load-time
resolver always decrypts with the default key, so encrypting with a non-default
key produces values that can't be loaded back. Hide the flag (still functional)
and drop it from the docs/examples until the round-trip is wired up.
`varlock load --format json-full` emits raw resolved secret values, so it must not go into an agent transcript. Recommend `--agent --format json-full` (redacted full graph) in the AI Tools guide, and flag the raw-secret caveat in the CLI reference.
--path/--env/--clear-cache/--skip-cache were repeated across 8 commands and had drifted (reveal's --path lost its multi-path note). Document them once in a Common options section; each command now references them and calls out only real deviations (scan/audit use --path as the schema entry point; audit is single-path).
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Updates gunshi
0.32→0.35.1, adopts two new features from it, and does a docs/--helppass focused on agent usefulness.gunshi upgrade
gunshiand@gunshi/plugin-completionto0.35.1, and pins@gunshi/plugin-i18n@^0.35.1(it was being auto-installed at a stale0.32.0, causing a peer-dep warning). No breaking changes in this range.varlock cache→ nested subcommandsstatusandclearare now real gunshi subcommands instead of hand-parsed positionals.--plugin/--yesare scoped toclear, each subcommand gets its own--help, and they show up in shell completion. Barevarlock cachestill opens the interactive browser.Positional arguments
printenv,explain,reveal,scan,auditnow declare their positional args (optional /multiple) instead of slicingctx.positionals. They render anARGUMENTS:section in--helpand are exposed to completion. Same user-facing syntax as before.Hidden args
run --no-inject-graph(deprecated) is markedhidden— still functional for back-compat, just removed from--help. Docs now document the current--inject <all|vars|blob>flag instead.encrypt --key-idis markedhidden— multi-key isn't round-trippable yet (thevarlock("local:...")reference doesn't encode a keyId and the load-time resolver always decrypts with the default key), so it stays functional but unadvertised until that's wired up.CLI docs &
--helpreview (agent lens)explainandinstall-plugincommands in the CLI reference.load/run/printenv/scan(audit already had them).loadJSON output contract (json/--agent/json-full) and the stdout/stderr summary flags for machine consumers; cross-linked these in the AI Tools guide.encrypt: added EXAMPLES (incl. stdin usage) and fixed a doubled default in--key-idhelp.Tests
scan/audit/explainpositionals and a--no-inject-graphback-compat guard.captureOutputoption and corrected the README (it's a pnpm workspace;bun testdoesn't apply).The
smoke-testslabel is applied so the smoke suite runs on this PR.