0.1c — amico-validate CLI + CI fast-tier schema gate (closes #17) by Rchari1 · Pull Request #30 · harmoniqs/amicode

Rchari1 · 2026-06-27T04:52:09Z

Stacked on #29 (0.1b). Do not merge.

A user-facing validator over the shared @amicode/schema validate() core, plus a CI step that reds the PR on a schema-invalid config — no parallel validation logic.

What's here

src/cli.ts: amico-validate <file> [--schema <kind>]. File→schema by basename for the fixed-filename artifacts (manifest.toml/result.toml/lab.toml/FINISHED via kindForFilename); --schema override for solvespec/catalog-entry or any non-standard name. Exit 0 valid / 64 invalid-or-usage (Q85). Field-precise stderr (offending key + JSON-pointer path).
bin amico-validate → dist/amico-validate.js (esbuild, shebang + chmod; ajv bundles).
test/fixtures/invalid/: one representative malformed file per kind — the shared negative corpus (the 0.1a valid corpus is the shared positive corpus; one corpus, all surfaces).
test/cli.test.ts (8): valid→0 + invalid→64 across every kind, file-role resolution, FINISHED-by-basename, non-filename-needs---schema, field-precise stderr, usage→64, --help→0.
ci.yml: fast-tier amico-validate gate over the shipped configs (lab.toml.example + demo manifest/result/FINISHED) after build.

Review

Adversarial code review: no must-fix / should-fix correctness. Verified empirically — every argv/edge path returns the right exit, kindForFilename doesn't misresolve (result.toml.bak → undefined), the CI bash word-split works (local zsh-only quirk), the gate runs after build, shebang + chmod +x present. One optional hardening nit (CI invokes the bundle directly rather than the linked bin) — non-blocking.

Tests

schema 32 / amico-run 47 / extension 63 green; CLI bundle 308kb.

🤖 Generated with Claude Code

jack-champagne

CLI's correct and the gate genuinely fails red — I checked the bash -e errexit path, a malformed config (even a middle one) reds the step. Two changes before merge, both small (inline). Plus the manifest.toml refs here must move in lockstep with #28's run.toml rename.

Tests + Phase 1: two test-debt items vs #17 — the invalid-fixture corpus is one-class-per-file with no wrong-type case (only an in-test mutation covers it; AC7 wants the class matrix self-contained), and the linked amico-validate bin is never exercised in CI (inline). The bigger Phase-1 item: this gate validates only static committed artifacts. Phase 1's reference-prompt smoke corpus IS the real emit→validate gate, and this CI step is where it plugs in — wire one real amico-run → emit run-dir → validate lane here (or in #31/#18), not just configs, so the producer seam is actually covered before Phase 1 piles E2E on top. That's also the infra decision: the cloud solve-service (aws-infra) writes this same run-dir contract to S3, so the gate the cloud track inherits should have seen real emitted output, not just fixtures.

Don't overbuild past this — one ci.yml edit + a build-order fix, not more machinery.

jack-champagne · 2026-06-27T21:34:31Z

+          $V packages/extension/scripts/lab.toml.example --schema lab
+          $V packages/extension/demo/run/manifest.toml
+          $V packages/extension/demo/run/result.toml
+          $V packages/extension/demo/run/FINISHED


[important] the gate covers lab/manifest/result/FINISHED but not solvespec or catalog-entry — AC5 requires the non-filename schemas via --schema, and the DoD is "all five." Both fixtures are committed; add them:

Suggested change

$V packages/extension/demo/run/FINISHED

$V packages/extension/demo/run/FINISHED

$V packages/schema/test/fixtures/valid/solvespec.toml --schema solvespec

$V packages/schema/test/fixtures/valid/catalog-entry.toml --schema catalog-entry

Done — the gate now also validates solvespec + catalog-entry via --schema (the committed valid fixtures), so all five are covered (AC5 / DoD).

jack-champagne · 2026-06-27T21:34:31Z

+        run: |
+          V="node packages/schema/dist/amico-validate.js"
+          $V packages/extension/scripts/lab.toml.example --schema lab
+          $V packages/extension/demo/run/manifest.toml


[rename] this manifest.toml becomes run.toml when #28 lands — update here in lockstep or the gate validates a filename that no longer exists.

Done — renamed to run.toml in lockstep with #28.

jack-champagne · 2026-06-27T21:34:31Z

  "type": "module",
  "main": "./src/index.ts",
  "types": "./src/index.ts",
+  "bin": { "amico-validate": "./dist/amico-validate.js" },


[important] this bin doesn't link on a clean install — dist/amico-validate.js is built by esbuild after pnpm install, so install warns Failed to create bin … ENOENT and the amico-validate command is never created. CI misses it because it runs the bundle path directly. For a slice whose deliverable is "ship the CLI," that's the entry point broken. Fix: a prepare script that builds on install, and have CI exercise the linked bin (pnpm exec amico-validate --help) so a broken entry point reds.

Fixed in this push — the bin points at a committed launcher (packages/schema/launcher/amico-validate, mirroring amico-run) that exists at install time, so the bin links cleanly (the dist bundle it execs is built by pnpm build). CI now runs pnpm --filter amicode-v2 exec amico-validate --help (the bin links into the dependents, not the root) so a broken entry point reds.

jack-champagne · 2026-06-27T21:34:31Z

+ *  --schema. The amico-validate CLI uses this for file-role resolution. */
+export function kindForFilename(filePath: string): SchemaKind | undefined {
+  const base = filePath.replace(/^.*[\\/]/, "");
+  if (base === "manifest.toml") return "manifest";


[rename] this manifest.toml (+ the SCHEMAS.manifest key, the schema filename, and the ci.yml gate line) must follow #28's run.toml rename together — miss one and basename resolution silently falls through to "cannot infer."

Done — kindForFilename now resolves run.toml→"run", and the SCHEMAS key / schema filename / ci.yml gate / cli.test KINDS all follow #28s rename. Added a self-contained wrong-type fixture so the class matrix no longer leans on an in-test mutation.

A user-facing validator over the shared @amicode/schema validate() core, plus a CI step that reds the PR on a schema-invalid config. No parallel validation logic. - src/cli.ts: `amico-validate <file> [--schema <kind>]`. Resolves file→schema by basename for the fixed-filename artifacts (manifest.toml/result.toml/lab.toml/ FINISHED via kindForFilename); --schema override for solvespec/catalog-entry or any non-standard name. Exit 0 valid / 64 invalid-or-usage (Q85). Field-precise stderr (offending key + JSON-pointer path). - bin amico-validate → dist/amico-validate.js (esbuild, shebang+chmod); ajv bundles. - test/fixtures/invalid/: one representative malformed file per kind (the shared negative corpus); valid corpus from 0.1a is the shared positive corpus. - test/cli.test.ts (8): valid→0 + invalid→64 across every kind, file-role resolution, FINISHED-by-basename, non-filename-needs-schema, field-precise stderr, usage errors → 64, --help → 0. - ci.yml: fast-tier `amico-validate` gate over the SHIPPED configs (lab.toml.example + demo/run manifest/result/FINISHED) after build. schema 32 / amico-run 47 / extension 63 green; CLI bundle 308kb. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

…all five, wrong-type fixture, rename lockstep Address Jack's #30 review (rebased on #28's run.toml rename): - bin now points at a COMMITTED launcher (packages/schema/launcher/amico-validate, mirrors amico-run) instead of the un-built dist/amico-validate.js — so the `amico-validate` bin links on a clean install (was "Failed to create bin … ENOENT"; dist is built by `pnpm build`, absent at install time). CI now exercises the LINKED bin (`pnpm --filter amicode-v2 exec amico-validate --help`) so a broken entry reds. - CI gate validates ALL FIVE: added solvespec + catalog-entry via --schema (AC5 / DoD). - Self-contained wrong-type fixture (invalid/result-wrongtype.toml) + a cli.test case — the class matrix no longer relies on an in-test mutation (AC7). - Rename lockstep: index.ts kindForFilename + cli.ts usage + ci.yml gate + cli.test KINDS + invalid/run.toml all follow #28's manifest.toml→run.toml / "manifest"→"run". schema 34 / amico-run 47 / extension 64 (+1 skip) green; linked bin + gate verified locally. (The bigger real-emit→validate lane Jack flagged lands in #31, the round-trip slice.) Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

…nd-to-end (rebased on #28 rename) Address Jack's #31 review: - REAL producer lane (not just fixtures): scripts/producer_roundtrip.sh runs amico-run on a minimal non-Piccolo emitter (julia/emit_fixture.jl) so it produces a live run-dir (run.toml + FINISHED from the orchestrator, result.toml from the Julia emitter), then validates the FRESHLY-EMITTED artifacts with validate.jl. CI runs it (fast tier; the real Piccolo solve stays the slow/nightly extension). Scoped to what each side actually emits — the orchestrator writes run.toml/FINISHED, the Julia producer writes result.toml. - AC6 exercised, not asserted: scripts/ac6_drift_check.sh perturbs the SHARED schema, REBUILDS TS (it bakes schemas at build time — the gap Jack found), then asserts BOTH the TS bin AND validate.jl reject a previously-valid fixture, then reverts. Proven locally; wired into the schema-roundtrip CI job. runtests.jl's AC6 testset notes the build-vs- runtime asymmetry. - Rename lockstep: validate.jl + runtests.jl manifest.toml→run.toml, "manifest"→"run", run.schema.json; resolved the ci.yml conflict keeping #30's all-five gate + this job. schema 34 / amico-run 47 / extension 64 (+1 skip) / julia 22 green; producer + AC6 scripts pass locally. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

jack-champagne

Re-reviewed c8f9863a. All three resolved:

Wrong-type fixture self-contained — result-wrongtype.toml (fidelity="high", a string where a float is required); AC7's class matrix no longer leans on an in-test mutation.
Linked bin exercised + gating in CI — the fast job runs pnpm --filter amicode-v2 exec amico-validate --help (a broken bin entry reds) and the committed launcher validates all five schemas under bash -e.
manifest.toml → run.toml in lockstep with #28.

Approving. Retargeting to main + merging behind #29.

Rchari1 mentioned this pull request Jun 27, 2026

0.1d — Julia round-trip: validate the emitted result.toml + run-dir against the shared schemas (closes #18) #31

Merged

jack-champagne requested changes Jun 27, 2026

View reviewed changes

Rchari1 force-pushed the rchari/0.1b-lab-validation branch from 5def952 to ade5a6d Compare June 28, 2026 20:29

Rchari1 and others added 2 commits June 28, 2026 16:29

Rchari1 force-pushed the rchari/0.1c-amico-validate-cli branch from f8daf3f to c8f9863 Compare June 28, 2026 20:35

Rchari1 requested a review from jack-champagne June 28, 2026 20:35

jack-champagne approved these changes Jun 29, 2026

View reviewed changes

jack-champagne changed the base branch from rchari/0.1b-lab-validation to main June 29, 2026 03:48

jack-champagne merged commit 2f9ec39 into main Jun 29, 2026
3 checks passed

jack-champagne mentioned this pull request Jun 30, 2026

0.2c — Route per-iter PNGs through LivePulsePlotCallback (AbstractIntermediateCallback) #43

Merged

kateebonner self-assigned this Jul 1, 2026

aarontrowbridge mentioned this pull request Jul 9, 2026

B6: one-source skills/agents + cross-runtime compiler #110

Open

3 tasks

Navigation Menu

Uh oh!

0.1c — amico-validate CLI + CI fast-tier schema gate (closes #17) - #30

0.1c — amico-validate CLI + CI fast-tier schema gate (closes #17)#30
jack-champagne merged 2 commits into
mainfrom
rchari/0.1c-amico-validate-cli

Rchari1 commented Jun 27, 2026

Uh oh!

jack-champagne left a comment

Uh oh!

jack-champagne Jun 27, 2026

Uh oh!

Rchari1 Jun 28, 2026

Uh oh!

jack-champagne Jun 27, 2026

Uh oh!

Rchari1 Jun 28, 2026

Uh oh!

jack-champagne Jun 27, 2026

Uh oh!

Rchari1 Jun 28, 2026

Uh oh!

jack-champagne Jun 27, 2026

Uh oh!

Rchari1 Jun 28, 2026

Uh oh!

jack-champagne left a comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Conversation

Rchari1 commented Jun 27, 2026

What's here

Review

Tests

Uh oh!

jack-champagne left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jack-champagne left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants