From 8af68207f51b0c03775be291cdd0700d3613f6fc Mon Sep 17 00:00:00 2001 From: Garry Tan Date: Sun, 19 Apr 2026 09:02:13 +0800 Subject: [PATCH] =?UTF-8?q?chore:=20remove=20gstack-publish=20=E2=80=94=20?= =?UTF-8?q?no=20real=20user=20need?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit User feedback: "i don't think i would use gstack-publish, i think we should remove it." Agreed. The CLI + marketplace wiring was an ambitious but speculative primitive. Zero users, zero validated demand, and the existing manual `clawhub publish` workflow already covers the real case (OpenClaw methodology skill publishing). Deleted: - bin/gstack-publish (the CLI) - skills.json (the marketplace manifest) - test/publish-dry-run.test.ts (13 tests) - ship/SKILL.md.tmpl Step 19.5 — the methodology-skill publish-on-ship check. No target to dispatch to anymore. - README.md Power tools row for gstack-publish Updated: - bin/gstack-model-benchmark doc comment: dropped "matches gstack-publish --dry-run semantics" reference (self-describing flag now) - CHANGELOG 1.3.0.0 entry: * Release summary: "three new binaries" → "two new binaries". Dropped the /ship publish-check narrative. * Numbers table: "1 of 3 → 3 of 3 wired" → "1 of 2 → 2 of 2 wired". Deterministic test count: 45 → 32 (removed publish-dry-run's 13). * Added section: removed gstack-publish CLI bullet + /ship Step 19.5 bullet. * "What this means for users" closer: replaced the /ship publish paragraph with the design-taste-engine learning loop, which IS real, wired, and something users hit every week via /design-shotgun. * Contributors section: "Four new test files" → "Three new test files" Retained: - openclaw/skills/gstack-openclaw-* skill dirs (pre-existed this PR, still publishable manually via `clawhub publish`, useful standalone for ClawHub installs) - CLAUDE.md publishing-native-skills section (same rationale) Regenerated SKILL.md across all hosts. Ship golden fixtures refreshed for claude/codex/factory. 455 tests pass. Co-Authored-By: Claude Opus 4.7 (1M context) --- CHANGELOG.md | 16 +- README.md | 1 - bin/gstack-model-benchmark | 1 - bin/gstack-publish | 264 ------------------- ship/SKILL.md | 33 +-- ship/SKILL.md.tmpl | 33 +-- skills.json | 90 ------- test/fixtures/golden/claude-ship-SKILL.md | 33 +-- test/fixtures/golden/codex-ship-SKILL.md | 33 +-- test/fixtures/golden/factory-ship-SKILL.md | 33 +-- test/publish-dry-run.test.ts | 280 --------------------- 11 files changed, 12 insertions(+), 805 deletions(-) delete mode 100755 bin/gstack-publish delete mode 100644 skills.json delete mode 100644 test/publish-dry-run.test.ts diff --git a/CHANGELOG.md b/CHANGELOG.md index 53a9e580..d6bedcde 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -5,16 +5,16 @@ ## **Every new CLI wired to a slash command.** ## **Zero orphan binaries ship in v1.3.** -v1.3 ships three new binaries (`gstack-model-benchmark`, `gstack-publish`, `gstack-taste-update`), one new skill (`/benchmark-models`), and a `/ship` Step 19.5 that detects methodology-skill changes and offers to publish them. The delta from v1.2 isn't just "more features." It's that every new primitive is discoverable from a `/command`, not buried in a CHANGELOG bullet nobody reads. Before this cut, `gstack-model-benchmark` and `gstack-publish` existed but no skill called them, so most users would never find them. Now `/benchmark-models` walks you through a cross-model comparison, and `/ship` asks about publishing the moment you touch a methodology skill. First multi-provider benchmark in any agent framework, and it's one slash command away. +v1.3 ships two new binaries (`gstack-model-benchmark`, `gstack-taste-update`) and one new skill (`/benchmark-models`). The delta from v1.2 isn't just "more features." It's that every new primitive is discoverable from a `/command`, not buried in a CHANGELOG bullet nobody reads. Before this cut, `gstack-model-benchmark` existed but no skill called it, so most users would never find it. Now `/benchmark-models` walks you through a cross-model comparison, and `gstack-taste-update` is already wired into `/design-shotgun`'s approval flow. First multi-provider benchmark in any agent framework, and it's one slash command away. ### The numbers that matter -Headline from this branch's review audit against `origin/main` (32 commits, commit `09466734`). Reproducible: `git log origin/main..HEAD --oneline` for the commit set, `bun test test/taste-engine.test.ts test/publish-dry-run.test.ts test/benchmark-cli.test.ts test/skill-e2e-benchmark-providers.test.ts` for the test count, `grep -rn "gstack-model-benchmark\|gstack-publish\|gstack-taste-update" --include="*.tmpl"` for wiring status. +Headline from this branch's review audit against `origin/main` (33 commits). Reproducible: `git log origin/main..HEAD --oneline` for the commit set, `bun test test/taste-engine.test.ts test/benchmark-cli.test.ts test/skill-e2e-benchmark-providers.test.ts` for the test count, `grep -rn "gstack-model-benchmark\|gstack-taste-update" --include="*.tmpl"` for wiring status. | Metric | BEFORE (initial v1.2 scope) | AFTER (v1.3) | Δ | |--------------------------------------------------|------------------------------|----------------------|-------------| -| **New CLIs wired to a /skill** | 1 of 3 (33%) | **3 of 3 (100%)** | **+2** | -| **Deterministic tests for v1.3 CLIs** | 0 | **45** | **+45** | +| **New CLIs wired to a /skill** | 1 of 2 (50%) | **2 of 2 (100%)** | **+1** | +| **Deterministic tests for v1.3 CLIs** | 0 | **32** | **+32** | | **Live-API adapter E2E (gated on `EVALS=1`)** | 0 | **8** | **+8** | | **Real adapter bugs caught by new tests** | 0 | **1** (codex `--skip-git-repo-check`) | **+1** | | **Preamble composition root** | 740 lines | **~100 lines** | **-86%** | @@ -24,7 +24,7 @@ The single most striking number: the new E2E suite caught a real codex adapter b ### What this means for gstack users -If you're a YC founder or solo builder shipping methodology skills from one laptop, `/benchmark-models` answers "is my skill better on Opus, GPT-5.4, or Gemini" with a real benchmark table, not vibes. When you tweak `/office-hours` or `/plan-ceo-review` on a feature branch, `/ship` asks whether to push to ClawHub + SkillsMP + Vercel Skills.sh too, so methodology updates don't die on your main branch. Continuous checkpoint mode (opt-in, local by default) means you can close your laptop mid-refactor and `/context-restore` picks you up from a WIP commit with decisions and remaining work intact, not a stale notes file. Run `/gstack-upgrade` and try `/benchmark-models` on the skill you use most this week. +If you're a YC founder or solo builder shipping methodology skills from one laptop, `/benchmark-models` answers "is my skill better on Opus, GPT-5.4, or Gemini" with a real benchmark table, not vibes. The design taste engine remembers which fonts, colors, and aesthetics you approve in `/design-shotgun`, so next round's variants lean toward your actual taste instead of resetting to Inter every time. Continuous checkpoint mode (opt-in, local by default) means you can close your laptop mid-refactor and `/context-restore` picks you up from a WIP commit with decisions and remaining work intact, not a stale notes file. Run `/gstack-upgrade` and try `/benchmark-models` on the skill you use most this week. ### Itemized changes @@ -36,14 +36,12 @@ If you're a YC founder or solo builder shipping methodology skills from one lapt - **Feature discovery prompt after upgrade.** When `JUST_UPGRADED` fires, gstack now offers to enable new features once per user (per-feature marker files at `~/.gstack/.feature-prompted-{name}`). Skipped entirely in spawned sessions (OpenClaw orchestrator). No more silent features that never get discovered. - **Context health soft directive (T2+ skills).** During long-running skills (/qa, /investigate, /cso), gstack now nudges you to write periodic `[PROGRESS]` summaries. Self-monitoring during 50+ tool-call sessions. No fake thresholds — soft directive that the model self-applies. Progress reporting never mutates git state. - **`gstack-model-benchmark` CLI.** Run the same prompt across Claude, GPT, and Gemini, compare latency/tokens/cost/quality. Per-provider auth detection, pricing tables, tool-compatibility map, parallel execution, per-provider error isolation. Quality scoring via Anthropic SDK as the stable judge (`--judge` flag). Output as table, JSON, or markdown. The first multi-provider benchmark in any agent framework — every other tool guesses which model is best, gstack measures it. -- **`gstack-publish` CLI for marketplace distribution.** Publishes gstack standalone methodology skills (gstack-office-hours, gstack-ceo-review, gstack-investigate, gstack-retro) to ClawHub, SkillsMP, and Vercel Skills.sh. Supports `--dry-run` (validate manifest + auth without publishing), per-skill error isolation (one failure doesn't abort the batch), idempotent re-runs. New `skills.json` manifest at the repo root declares per-skill marketplace targets and metadata. - **Design taste engine.** Persistent cross-session taste profile at `~/.gstack/projects/$SLUG/taste-profile.json`. Tracks fonts, colors, layouts, and aesthetic directions you approve and reject across sessions. Confidence decays 5% per week. Design-consultation and design-shotgun now factor in your demonstrated preferences. Schema migration handles legacy approved.json. New `gstack-taste-update` CLI updates the profile after design-shotgun decisions. - **Anti-slop design constraints.** Design-consultation now asks "What's the one thing someone will remember?" as a forcing question. Phase 5 self-gate: "Would a human designer be embarrassed by this?" — discards and regenerates if yes. Anti-convergence directive in design-shotgun: each variant must use a different font, palette, and layout, or one of them failed. Space Grotesk added to the overused fonts list (it's the new "safe alternative to Inter" trap). system-ui-as-primary-font added to the AI slop blacklist. - **`gstack-config list` and `gstack-config defaults`** subcommands. `list` shows all config keys with their current value AND source (set/default). `defaults` shows just the defaults table. Fixes the prior gap where `get` returned empty for missing keys instead of falling back to the documented defaults. Telemetry default aligned: header and runtime both say `off` now (previously mismatched). - **`gstack-config checkpoint_mode` and `checkpoint_push` keys.** New config knobs for continuous checkpoint mode. Both default to safe values (`explicit` mode, no auto-push). - **New `/benchmark-models` skill.** Wraps `gstack-model-benchmark` in an interactive flow: pick a prompt (an existing SKILL.md, inline text, or file path), confirm providers (dry-run shows auth status per provider), decide on `--judge` (adds ~$0.05 for quality scoring), run, interpret. Trigger phrases: "compare models", "model shootout", "which model is best". Separate from `/benchmark` (which measures web page performance) — different surface, different domain. -- **`/ship` Step 19.5 — methodology skill publish check.** When a PR touches any standalone methodology skill (`openclaw/skills/gstack-*/SKILL.md`) or the `skills.json` manifest, `/ship` runs `gstack-publish --dry-run` after PR creation and asks whether to actually publish. Conditional — silent no-op on PRs that don't touch methodology skills. Eliminates the "shipped to main but never pushed to the marketplace" failure mode. -- **`gstack-model-benchmark --dry-run`.** Offline validation mode that matches `gstack-publish --dry-run` semantics. Validates the provider list, resolves per-adapter auth, echoes the resolved flag values, and exits without invoking any provider CLI. Zero-cost pre-flight for CI pipelines and for catching auth drift before starting a paid benchmark run. +- **`gstack-model-benchmark --dry-run`.** Offline validation mode. Validates the provider list, resolves per-adapter auth, echoes the resolved flag values, and exits without invoking any provider CLI. Zero-cost pre-flight for CI pipelines and for catching auth drift before starting a paid benchmark run. ### Changed @@ -64,7 +62,7 @@ If you're a YC founder or solo builder shipping methodology skills from one lapt - **`scripts/resolvers/preamble/`** — 18 single-purpose generators, 16-160 lines each. The composition root in `scripts/resolvers/preamble.ts` imports them and wires them into the tier-gated section list. - **Plan and reviews persisted.** Implementation followed `~/.claude/plans/declarative-riding-cook.md` which went through CEO review (SCOPE EXPANSION, 6 expansions accepted), DX review (POLISH, 5 gaps fixed), Eng review (4 architecture issues), and Codex review (11 brutal findings, all integrated and 2 prior decisions reversed). - **Mode-posture energy in Writing Style rules 2-4** (ported from main's v1.1.2.0). Rule 2 and rule 4 now cover three framings — pain reduction, capability unlocked, forcing-question pressure — so expansion, builder, and forcing-question skills keep their edge instead of collapsing into diagnostic-pain framing. Rule 3 adds an explicit exception for stacked forcing questions. Came in via the merge; sits on top of the submodule refactor already shipped in v1.3. -- **Lite E2E coverage for v1.3 primitives.** Four new test files fill the real coverage gaps flagged in initial review: `test/taste-engine.test.ts` (24 tests — schema shape, Laplace-smoothed confidence, 5%/week decay clamped at 0, multi-dimension extraction, case-insensitive first-casing-wins policy, session cap via seed-then-one-call, legacy profile migration, taste-drift conflict warning, malformed-JSON recovery), `test/publish-dry-run.test.ts` (13 tests — manifest parsing, missing source file detection, slug filter, per-marketplace auth isolation via fake marketplaces), `test/benchmark-cli.test.ts` (12 tests — CLI flag wiring, provider defaults, unknown-provider WARN path, NOT-READY branch regression catcher that strips auth env vars), `test/skill-e2e-benchmark-providers.test.ts` (8 periodic-tier live-API tests — trivial "echo ok" prompt through claude/codex/gemini adapters, assertions on parsed output + tokens + cost + timeout error codes + Promise.allSettled parallel isolation). +- **Lite E2E coverage for v1.3 primitives.** Three new test files fill the real coverage gaps flagged in initial review: `test/taste-engine.test.ts` (24 tests — schema shape, Laplace-smoothed confidence, 5%/week decay clamped at 0, multi-dimension extraction, case-insensitive first-casing-wins policy, session cap via seed-then-one-call, legacy profile migration, taste-drift conflict warning, malformed-JSON recovery), `test/benchmark-cli.test.ts` (12 tests — CLI flag wiring, provider defaults, unknown-provider WARN path, NOT-READY branch regression catcher that strips auth env vars), `test/skill-e2e-benchmark-providers.test.ts` (8 periodic-tier live-API tests — trivial "echo ok" prompt through claude/codex/gemini adapters, assertions on parsed output + tokens + cost + timeout error codes + Promise.allSettled parallel isolation). - **Ship golden fixtures for three hosts.** `test/fixtures/golden/{claude,codex,factory}-ship-SKILL.md` — byte-exact regression pins on the `/ship` generated output. The adversarial subagent pass during /review caught two real bugs before merge: Geist/GEIST casing policy in the taste engine was unpinned, and the live-E2E workdir was created at module load and never cleaned up. ## [1.1.3.0] - 2026-04-19 diff --git a/README.md b/README.md index 633ef535..de28bbc6 100644 --- a/README.md +++ b/README.md @@ -234,7 +234,6 @@ Beyond the slash-command skills, gstack ships standalone CLIs for workflows that | Command | What it does | |---------|-------------| | `gstack-model-benchmark` | **Cross-model benchmark** — run the same prompt through Claude, GPT (via Codex CLI), and Gemini; compare latency, tokens, cost, and (optionally) LLM-judge quality score. Auth detected per provider, unavailable providers skip cleanly. Output as table, JSON, or markdown. `--dry-run` validates flags + auth without spending API calls. | -| `gstack-publish` | **Marketplace distribution** — publishes standalone methodology skills (office-hours, ceo-review, investigate, retro) to ClawHub, SkillsMP, and Vercel Skills.sh. Manifest at `skills.json`. `--dry-run` validates everything without actually publishing. Per-skill, per-marketplace error isolation — one failure never aborts the batch. | | `gstack-taste-update` | **Design taste learning** — writes approvals and rejections from `/design-shotgun` into a persistent per-project taste profile. Decays 5%/week. Feeds back into future variant generation so the system learns what you actually pick. | ### Continuous checkpoint mode (opt-in, local by default) diff --git a/bin/gstack-model-benchmark b/bin/gstack-model-benchmark index 301ddeea..7c48c910 100755 --- a/bin/gstack-model-benchmark +++ b/bin/gstack-model-benchmark @@ -17,7 +17,6 @@ * --judge Run Anthropic SDK judge on outputs for quality score * (requires ANTHROPIC_API_KEY; adds ~$0.05 per call) * --dry-run Validate flags + resolve auth, don't invoke providers - * (matches gstack-publish --dry-run semantics) * * Examples: * gstack-model-benchmark --prompt "Write a haiku about databases" --models claude,gpt diff --git a/bin/gstack-publish b/bin/gstack-publish deleted file mode 100755 index 0f040939..00000000 --- a/bin/gstack-publish +++ /dev/null @@ -1,264 +0,0 @@ -#!/usr/bin/env bun -/** - * gstack-publish — publish gstack standalone methodology skills to - * configured marketplaces (ClawHub, SkillsMP, Vercel Skills.sh). - * - * Usage: - * gstack-publish Publish all skills listed in skills.json - * gstack-publish Publish a single skill by slug - * gstack-publish --dry-run Validate manifest + resolve auth, don't publish - * gstack-publish --list List skills + marketplace targets - * gstack-publish --changelog "description" - * Pass a changelog message (required by some CLIs) - * - * Per-skill error isolation: if publishing to one marketplace fails for one skill, - * the batch continues. Aggregate failures are reported at the end with exit 1. - * - * Idempotent: publishing the same version twice is a no-op for marketplaces that - * reject duplicate versions; for others, gstack-publish bumps the version check - * via their auth_check command and reports if the version already exists. - */ - -import * as fs from 'fs'; -import * as path from 'path'; -import { execSync, spawnSync } from 'child_process'; - -const ROOT = path.resolve(import.meta.dir, '..'); -const MANIFEST_PATH = path.join(ROOT, 'skills.json'); - -interface SkillEntry { - slug: string; - source: string; - name: string; - version: string; - category: string; - description: string; - marketplaces: Record; - standalone: boolean; - compatible_hosts: string[]; -} - -interface MarketplaceConfig { - cli: string; - login_cmd: string; - publish_cmd_template: string; - docs: string; - auth_check: string; -} - -interface Manifest { - version: string; - description: string; - skills: SkillEntry[]; - marketplaces: Record; -} - -function arg(name: string, def?: string): string | undefined { - const idx = process.argv.findIndex(a => a === name || a.startsWith(name + '=')); - if (idx < 0) return def; - const eqIdx = process.argv[idx].indexOf('='); - if (eqIdx >= 0) return process.argv[idx].slice(eqIdx + 1); - return process.argv[idx + 1]; -} - -function flag(name: string): boolean { - return process.argv.includes(name); -} - -function loadManifest(): Manifest { - if (!fs.existsSync(MANIFEST_PATH)) { - console.error(`ERROR: ${MANIFEST_PATH} not found.`); - process.exit(1); - } - try { - return JSON.parse(fs.readFileSync(MANIFEST_PATH, 'utf-8')); - } catch (err) { - console.error(`ERROR: could not parse ${MANIFEST_PATH}:`, (err as Error).message); - process.exit(1); - } -} - -function validateSkill(skill: SkillEntry): string[] { - const errs: string[] = []; - if (!skill.slug) errs.push('missing slug'); - if (!skill.version) errs.push('missing version'); - if (!skill.source) errs.push('missing source'); - if (skill.source) { - const fullPath = path.join(ROOT, skill.source); - if (!fs.existsSync(fullPath)) errs.push(`source file missing: ${skill.source}`); - } - if (!Object.keys(skill.marketplaces || {}).length) errs.push('no marketplaces configured'); - return errs; -} - -function checkAuth(marketplace: string, cfg: MarketplaceConfig): { ok: boolean; reason?: string } { - // Check binary exists - const binCheck = spawnSync('sh', ['-c', `command -v ${cfg.cli}`], { timeout: 2000 }); - if (binCheck.status !== 0) { - return { ok: false, reason: `${cfg.cli} CLI not on PATH. Docs: ${cfg.docs}` }; - } - // Run auth check. Success = ok. Any non-zero = not authenticated. - const authCheck = spawnSync('sh', ['-c', cfg.auth_check], { timeout: 5000 }); - if (authCheck.status !== 0) { - return { ok: false, reason: `not authenticated. Run: ${cfg.login_cmd}` }; - } - return { ok: true }; -} - -interface PublishResult { - skill: string; - marketplace: string; - ok: boolean; - reason?: string; - cmd?: string; -} - -function buildCommand(template: string, subs: Record): string { - return template.replace(/\{(\w+)\}/g, (_, key) => subs[key] ?? `{${key}}`); -} - -function publishOne(skill: SkillEntry, marketplace: string, cfg: MarketplaceConfig, changelog: string, dryRun: boolean): PublishResult { - const sourceDir = path.dirname(path.join(ROOT, skill.source)); - const target = skill.marketplaces[marketplace]; - if (!target || !target.publish) { - return { skill: skill.slug, marketplace, ok: true, reason: 'not configured for this marketplace' }; - } - const cmd = buildCommand(cfg.publish_cmd_template, { - source_dir: sourceDir, - slug: target.slug, - name: skill.name, - version: skill.version, - changelog, - }); - if (dryRun) { - return { skill: skill.slug, marketplace, ok: true, reason: 'DRY-RUN (not executed)', cmd }; - } - try { - const out = execSync(cmd, { stdio: 'pipe', timeout: 120_000, encoding: 'utf-8' }); - return { skill: skill.slug, marketplace, ok: true, cmd, reason: out.slice(0, 200) }; - } catch (err: unknown) { - const e = err as { stderr?: Buffer; message?: string }; - const reason = e.stderr?.toString()?.slice(0, 400) ?? e.message ?? 'unknown error'; - return { skill: skill.slug, marketplace, ok: false, cmd, reason }; - } -} - -function cmdList(m: Manifest): void { - console.log(`gstack-publish manifest v${m.version}: ${m.skills.length} skills\n`); - for (const s of m.skills) { - console.log(` ${s.slug} v${s.version} (${s.category})`); - console.log(` source: ${s.source}`); - const targets = Object.entries(s.marketplaces).filter(([, v]) => v.publish).map(([k]) => k); - console.log(` publish to: ${targets.join(', ') || '(none)'}`); - console.log(); - } - console.log('Configured marketplaces:'); - for (const [name, cfg] of Object.entries(m.marketplaces)) { - console.log(` ${name}: ${cfg.cli} (docs: ${cfg.docs})`); - } -} - -async function main(): Promise { - const manifest = loadManifest(); - - if (flag('--list')) { - cmdList(manifest); - return; - } - - const dryRun = flag('--dry-run'); - const changelog = arg('--changelog') ?? `gstack-publish ${new Date().toISOString().slice(0, 10)}`; - const targetSlug = process.argv.slice(2).find(a => !a.startsWith('--')); - - // 1. Validate manifest entries - console.log(`== Validating manifest (${manifest.skills.length} skills) ==`); - const validationErrors: Array<{ slug: string; errors: string[] }> = []; - for (const s of manifest.skills) { - const errs = validateSkill(s); - if (errs.length) validationErrors.push({ slug: s.slug, errors: errs }); - } - if (validationErrors.length) { - console.error('Manifest validation failed:'); - for (const { slug, errors } of validationErrors) { - console.error(` ${slug}: ${errors.join(', ')}`); - } - process.exit(1); - } - console.log(' OK\n'); - - // 2. Filter skills (single slug or all) - const skills = targetSlug - ? manifest.skills.filter(s => s.slug === targetSlug) - : manifest.skills; - if (!skills.length) { - console.error(`ERROR: skill not found: ${targetSlug}`); - process.exit(1); - } - - // 3. Check marketplace auth (only for marketplaces we'll actually use) - const activeMarketplaces = new Set(); - for (const s of skills) { - for (const [name, cfg] of Object.entries(s.marketplaces)) { - if (cfg.publish) activeMarketplaces.add(name); - } - } - console.log(`== Checking auth for ${activeMarketplaces.size} marketplace(s) ==`); - const authStatus: Record = {}; - for (const mp of activeMarketplaces) { - const cfg = manifest.marketplaces[mp]; - if (!cfg) { - authStatus[mp] = { ok: false, reason: `no config for marketplace ${mp}` }; - continue; - } - const res = checkAuth(mp, cfg); - authStatus[mp] = res; - console.log(` ${mp}: ${res.ok ? 'OK' : 'NOT READY — ' + res.reason}`); - } - console.log(); - - // 4. Publish - const results: PublishResult[] = []; - for (const skill of skills) { - console.log(`== Publishing ${skill.slug} v${skill.version} ==`); - for (const [mp, target] of Object.entries(skill.marketplaces)) { - if (!target.publish) continue; - if (!authStatus[mp]?.ok && !dryRun) { - results.push({ skill: skill.slug, marketplace: mp, ok: false, reason: `auth not ready: ${authStatus[mp]?.reason}` }); - console.log(` ${mp}: SKIPPED (auth not ready)`); - continue; - } - const cfg = manifest.marketplaces[mp]; - if (!cfg) { - results.push({ skill: skill.slug, marketplace: mp, ok: false, reason: 'no marketplace config' }); - continue; - } - const r = publishOne(skill, mp, cfg, changelog, dryRun); - results.push(r); - const tag = r.ok ? (dryRun ? 'DRY-RUN' : 'OK') : 'FAIL'; - console.log(` ${mp}: ${tag}${r.reason ? ' — ' + r.reason.slice(0, 80) : ''}`); - } - console.log(); - } - - // 5. Summary - const successes = results.filter(r => r.ok).length; - const failures = results.filter(r => !r.ok); - console.log(`== Summary ==`); - console.log(` Published: ${successes}`); - console.log(` Failed: ${failures.length}`); - if (failures.length) { - console.log(); - for (const f of failures) { - console.log(` ${f.skill} → ${f.marketplace}: ${f.reason}`); - } - process.exit(1); - } - if (dryRun) { - console.log(`\n(--dry-run — no actual publish happened)`); - } -} - -main().catch(err => { - console.error('FATAL:', err); - process.exit(1); -}); diff --git a/ship/SKILL.md b/ship/SKILL.md index c67c931d..c0e14388 100644 --- a/ship/SKILL.md +++ b/ship/SKILL.md @@ -2930,38 +2930,7 @@ EOF **If neither CLI is available:** Print the branch name, remote URL, and instruct the user to create the PR/MR manually via the web UI. Do not stop — the code is pushed and ready. -**Output the PR/MR URL** — then proceed to Step 19.5. - ---- - -## Step 19.5: Offer methodology skill publishing (conditional) - -If this PR touched any standalone methodology skill (`openclaw/skills/gstack-*/SKILL.md`) or the marketplace manifest (`skills.json`), offer to publish the updated skills to the configured marketplaces after merge. - -```bash -git diff origin/...HEAD --name-only | grep -E '^(openclaw/skills/gstack-[^/]+/SKILL\.md|skills\.json)$' || true -``` - -If the output is empty: skip this step silently. Continue to Step 20. - -If any matches: run a dry-run preview so the user sees exactly what would publish and what auth is missing. - -```bash -~/.claude/skills/gstack/bin/gstack-publish --dry-run -``` - -Then use AskUserQuestion with the structure in the preamble: -- **Re-ground:** project, branch, PR just opened. -- **Simplify:** "This PR updated methodology skills. gstack-publish can push them to ClawHub, SkillsMP, and Vercel Skills.sh so other developers on other agents can install them. Dry-run above shows what would ship and whether each marketplace is authed." -- **RECOMMENDATION:** A because the dry-run just verified the manifest and one-command distribution is the whole point of having a marketplace publisher. -- **Options:** - - A) Publish now — run `gstack-publish` (human: ~2min / CC: ~30s). Completeness: 9/10. - - B) Publish after the PR merges — `/document-release` won't push this automatically; you'll run `gstack-publish` manually once the branch is on main. Completeness: 7/10. - - C) Skip — don't publish this release. Completeness: 4/10. - -If A: run `~/.claude/skills/gstack/bin/gstack-publish` (no --dry-run). Report success/failure per marketplace. If any marketplace auth is missing, the dry-run above surfaced it; the real publish will skip that marketplace with an isolated error rather than aborting the batch. - -If B or C: continue to Step 20. +**Output the PR/MR URL** — then proceed to Step 20. --- diff --git a/ship/SKILL.md.tmpl b/ship/SKILL.md.tmpl index c4f706a6..9eab6d33 100644 --- a/ship/SKILL.md.tmpl +++ b/ship/SKILL.md.tmpl @@ -854,38 +854,7 @@ EOF **If neither CLI is available:** Print the branch name, remote URL, and instruct the user to create the PR/MR manually via the web UI. Do not stop — the code is pushed and ready. -**Output the PR/MR URL** — then proceed to Step 19.5. - ---- - -## Step 19.5: Offer methodology skill publishing (conditional) - -If this PR touched any standalone methodology skill (`openclaw/skills/gstack-*/SKILL.md`) or the marketplace manifest (`skills.json`), offer to publish the updated skills to the configured marketplaces after merge. - -```bash -git diff origin/...HEAD --name-only | grep -E '^(openclaw/skills/gstack-[^/]+/SKILL\.md|skills\.json)$' || true -``` - -If the output is empty: skip this step silently. Continue to Step 20. - -If any matches: run a dry-run preview so the user sees exactly what would publish and what auth is missing. - -```bash -~/.claude/skills/gstack/bin/gstack-publish --dry-run -``` - -Then use AskUserQuestion with the structure in the preamble: -- **Re-ground:** project, branch, PR just opened. -- **Simplify:** "This PR updated methodology skills. gstack-publish can push them to ClawHub, SkillsMP, and Vercel Skills.sh so other developers on other agents can install them. Dry-run above shows what would ship and whether each marketplace is authed." -- **RECOMMENDATION:** A because the dry-run just verified the manifest and one-command distribution is the whole point of having a marketplace publisher. -- **Options:** - - A) Publish now — run `gstack-publish` (human: ~2min / CC: ~30s). Completeness: 9/10. - - B) Publish after the PR merges — `/document-release` won't push this automatically; you'll run `gstack-publish` manually once the branch is on main. Completeness: 7/10. - - C) Skip — don't publish this release. Completeness: 4/10. - -If A: run `~/.claude/skills/gstack/bin/gstack-publish` (no --dry-run). Report success/failure per marketplace. If any marketplace auth is missing, the dry-run above surfaced it; the real publish will skip that marketplace with an isolated error rather than aborting the batch. - -If B or C: continue to Step 20. +**Output the PR/MR URL** — then proceed to Step 20. --- diff --git a/skills.json b/skills.json deleted file mode 100644 index eef3dc46..00000000 --- a/skills.json +++ /dev/null @@ -1,90 +0,0 @@ -{ - "$schema": "https://gstack.dev/schemas/skills-manifest.v1.json", - "version": "1.0.0", - "description": "Manifest for publishing gstack standalone methodology skills to marketplaces (ClawHub, Vercel Skills.sh, SkillsMP). These skills have NO gstack runtime dependencies — they work as pure Claude Code / OpenClaw skills.", - "skills": [ - { - "slug": "gstack-office-hours", - "source": "openclaw/skills/gstack-openclaw-office-hours/SKILL.md", - "name": "YC Office Hours", - "version": "1.0.0", - "category": "strategy", - "description": "Product interrogation with six forcing questions. Two modes: startup diagnostic and builder brainstorm. Produces a design doc, not code.", - "marketplaces": { - "clawhub": { "slug": "gstack-office-hours", "publish": true }, - "skillsmp": { "slug": "gstack-office-hours", "publish": true }, - "vercel": { "slug": "gstack-office-hours", "publish": true } - }, - "standalone": true, - "compatible_hosts": ["claude-code", "openclaw", "cursor"] - }, - { - "slug": "gstack-ceo-review", - "source": "openclaw/skills/gstack-openclaw-ceo-review/SKILL.md", - "name": "CEO Plan Review", - "version": "1.0.0", - "category": "strategy", - "description": "Founder-mode plan review. Rethink the problem, find the 10x product, challenge premises, expand scope when it creates a better product.", - "marketplaces": { - "clawhub": { "slug": "gstack-ceo-review", "publish": true }, - "skillsmp": { "slug": "gstack-ceo-review", "publish": true }, - "vercel": { "slug": "gstack-ceo-review", "publish": true } - }, - "standalone": true, - "compatible_hosts": ["claude-code", "openclaw", "cursor"] - }, - { - "slug": "gstack-investigate", - "source": "openclaw/skills/gstack-openclaw-investigate/SKILL.md", - "name": "Root Cause Investigation", - "version": "1.0.0", - "category": "debugging", - "description": "Systematic debugging with root cause investigation. Four phases: investigate, analyze, hypothesize, implement. Iron Law: no fixes without root cause.", - "marketplaces": { - "clawhub": { "slug": "gstack-investigate", "publish": true }, - "skillsmp": { "slug": "gstack-investigate", "publish": true }, - "vercel": { "slug": "gstack-investigate", "publish": true } - }, - "standalone": true, - "compatible_hosts": ["claude-code", "openclaw", "cursor"] - }, - { - "slug": "gstack-retro", - "source": "openclaw/skills/gstack-openclaw-retro/SKILL.md", - "name": "Engineering Retrospective", - "version": "1.0.0", - "category": "process", - "description": "Weekly engineering retrospective. Analyzes commit history and work patterns with per-person breakdown.", - "marketplaces": { - "clawhub": { "slug": "gstack-retro", "publish": true }, - "skillsmp": { "slug": "gstack-retro", "publish": true }, - "vercel": { "slug": "gstack-retro", "publish": true } - }, - "standalone": true, - "compatible_hosts": ["claude-code", "openclaw", "cursor"] - } - ], - "marketplaces": { - "clawhub": { - "cli": "clawhub", - "login_cmd": "clawhub login", - "publish_cmd_template": "clawhub publish {source_dir} --slug {slug} --name \"{name}\" --version {version} --changelog \"{changelog}\"", - "docs": "https://clawhub.dev/docs/publishing", - "auth_check": "clawhub whoami" - }, - "skillsmp": { - "cli": "skillsmp", - "login_cmd": "skillsmp login", - "publish_cmd_template": "skillsmp publish {source_dir} --name {slug} --version {version}", - "docs": "https://skillsmp.com/docs/publish", - "auth_check": "skillsmp whoami" - }, - "vercel": { - "cli": "skills", - "login_cmd": "skills login", - "publish_cmd_template": "skills publish {source_dir} --package {slug} --version {version}", - "docs": "https://skills.sh/docs", - "auth_check": "skills whoami" - } - } -} diff --git a/test/fixtures/golden/claude-ship-SKILL.md b/test/fixtures/golden/claude-ship-SKILL.md index c67c931d..c0e14388 100644 --- a/test/fixtures/golden/claude-ship-SKILL.md +++ b/test/fixtures/golden/claude-ship-SKILL.md @@ -2930,38 +2930,7 @@ EOF **If neither CLI is available:** Print the branch name, remote URL, and instruct the user to create the PR/MR manually via the web UI. Do not stop — the code is pushed and ready. -**Output the PR/MR URL** — then proceed to Step 19.5. - ---- - -## Step 19.5: Offer methodology skill publishing (conditional) - -If this PR touched any standalone methodology skill (`openclaw/skills/gstack-*/SKILL.md`) or the marketplace manifest (`skills.json`), offer to publish the updated skills to the configured marketplaces after merge. - -```bash -git diff origin/...HEAD --name-only | grep -E '^(openclaw/skills/gstack-[^/]+/SKILL\.md|skills\.json)$' || true -``` - -If the output is empty: skip this step silently. Continue to Step 20. - -If any matches: run a dry-run preview so the user sees exactly what would publish and what auth is missing. - -```bash -~/.claude/skills/gstack/bin/gstack-publish --dry-run -``` - -Then use AskUserQuestion with the structure in the preamble: -- **Re-ground:** project, branch, PR just opened. -- **Simplify:** "This PR updated methodology skills. gstack-publish can push them to ClawHub, SkillsMP, and Vercel Skills.sh so other developers on other agents can install them. Dry-run above shows what would ship and whether each marketplace is authed." -- **RECOMMENDATION:** A because the dry-run just verified the manifest and one-command distribution is the whole point of having a marketplace publisher. -- **Options:** - - A) Publish now — run `gstack-publish` (human: ~2min / CC: ~30s). Completeness: 9/10. - - B) Publish after the PR merges — `/document-release` won't push this automatically; you'll run `gstack-publish` manually once the branch is on main. Completeness: 7/10. - - C) Skip — don't publish this release. Completeness: 4/10. - -If A: run `~/.claude/skills/gstack/bin/gstack-publish` (no --dry-run). Report success/failure per marketplace. If any marketplace auth is missing, the dry-run above surfaced it; the real publish will skip that marketplace with an isolated error rather than aborting the batch. - -If B or C: continue to Step 20. +**Output the PR/MR URL** — then proceed to Step 20. --- diff --git a/test/fixtures/golden/codex-ship-SKILL.md b/test/fixtures/golden/codex-ship-SKILL.md index dc523a48..cfa85e6e 100644 --- a/test/fixtures/golden/codex-ship-SKILL.md +++ b/test/fixtures/golden/codex-ship-SKILL.md @@ -2545,38 +2545,7 @@ EOF **If neither CLI is available:** Print the branch name, remote URL, and instruct the user to create the PR/MR manually via the web UI. Do not stop — the code is pushed and ready. -**Output the PR/MR URL** — then proceed to Step 19.5. - ---- - -## Step 19.5: Offer methodology skill publishing (conditional) - -If this PR touched any standalone methodology skill (`openclaw/skills/gstack-*/SKILL.md`) or the marketplace manifest (`skills.json`), offer to publish the updated skills to the configured marketplaces after merge. - -```bash -git diff origin/...HEAD --name-only | grep -E '^(openclaw/skills/gstack-[^/]+/SKILL\.md|skills\.json)$' || true -``` - -If the output is empty: skip this step silently. Continue to Step 20. - -If any matches: run a dry-run preview so the user sees exactly what would publish and what auth is missing. - -```bash -$GSTACK_ROOT/bin/gstack-publish --dry-run -``` - -Then use AskUserQuestion with the structure in the preamble: -- **Re-ground:** project, branch, PR just opened. -- **Simplify:** "This PR updated methodology skills. gstack-publish can push them to ClawHub, SkillsMP, and Vercel Skills.sh so other developers on other agents can install them. Dry-run above shows what would ship and whether each marketplace is authed." -- **RECOMMENDATION:** A because the dry-run just verified the manifest and one-command distribution is the whole point of having a marketplace publisher. -- **Options:** - - A) Publish now — run `gstack-publish` (human: ~2min / CC: ~30s). Completeness: 9/10. - - B) Publish after the PR merges — `/document-release` won't push this automatically; you'll run `gstack-publish` manually once the branch is on main. Completeness: 7/10. - - C) Skip — don't publish this release. Completeness: 4/10. - -If A: run `$GSTACK_ROOT/bin/gstack-publish` (no --dry-run). Report success/failure per marketplace. If any marketplace auth is missing, the dry-run above surfaced it; the real publish will skip that marketplace with an isolated error rather than aborting the batch. - -If B or C: continue to Step 20. +**Output the PR/MR URL** — then proceed to Step 20. --- diff --git a/test/fixtures/golden/factory-ship-SKILL.md b/test/fixtures/golden/factory-ship-SKILL.md index 9e8d4745..cba656ba 100644 --- a/test/fixtures/golden/factory-ship-SKILL.md +++ b/test/fixtures/golden/factory-ship-SKILL.md @@ -2921,38 +2921,7 @@ EOF **If neither CLI is available:** Print the branch name, remote URL, and instruct the user to create the PR/MR manually via the web UI. Do not stop — the code is pushed and ready. -**Output the PR/MR URL** — then proceed to Step 19.5. - ---- - -## Step 19.5: Offer methodology skill publishing (conditional) - -If this PR touched any standalone methodology skill (`openclaw/skills/gstack-*/SKILL.md`) or the marketplace manifest (`skills.json`), offer to publish the updated skills to the configured marketplaces after merge. - -```bash -git diff origin/...HEAD --name-only | grep -E '^(openclaw/skills/gstack-[^/]+/SKILL\.md|skills\.json)$' || true -``` - -If the output is empty: skip this step silently. Continue to Step 20. - -If any matches: run a dry-run preview so the user sees exactly what would publish and what auth is missing. - -```bash -$GSTACK_ROOT/bin/gstack-publish --dry-run -``` - -Then use AskUserQuestion with the structure in the preamble: -- **Re-ground:** project, branch, PR just opened. -- **Simplify:** "This PR updated methodology skills. gstack-publish can push them to ClawHub, SkillsMP, and Vercel Skills.sh so other developers on other agents can install them. Dry-run above shows what would ship and whether each marketplace is authed." -- **RECOMMENDATION:** A because the dry-run just verified the manifest and one-command distribution is the whole point of having a marketplace publisher. -- **Options:** - - A) Publish now — run `gstack-publish` (human: ~2min / CC: ~30s). Completeness: 9/10. - - B) Publish after the PR merges — `/document-release` won't push this automatically; you'll run `gstack-publish` manually once the branch is on main. Completeness: 7/10. - - C) Skip — don't publish this release. Completeness: 4/10. - -If A: run `$GSTACK_ROOT/bin/gstack-publish` (no --dry-run). Report success/failure per marketplace. If any marketplace auth is missing, the dry-run above surfaced it; the real publish will skip that marketplace with an isolated error rather than aborting the batch. - -If B or C: continue to Step 20. +**Output the PR/MR URL** — then proceed to Step 20. --- diff --git a/test/publish-dry-run.test.ts b/test/publish-dry-run.test.ts deleted file mode 100644 index 4088f447..00000000 --- a/test/publish-dry-run.test.ts +++ /dev/null @@ -1,280 +0,0 @@ -/** - * gstack-publish end-to-end tests via --dry-run. - * - * Verifies manifest parsing, schema validation, marketplace auth checks, per-skill - * error isolation, and command building — all without touching real marketplaces. - * - * --dry-run does NOT run execSync on publish commands. Auth checks still run - * against real binaries; we use fake marketplaces whose `auth_check` commands - * are always-succeed (`true`) or always-fail (`false`) so the test is hermetic. - */ - -import { describe, test, expect, beforeEach, afterEach } from 'bun:test'; -import { spawnSync } from 'child_process'; -import * as fs from 'fs'; -import * as path from 'path'; -import * as os from 'os'; - -const ROOT = path.resolve(import.meta.dir, '..'); -const BIN = path.join(ROOT, 'bin', 'gstack-publish'); - -let sandbox: string; -let binCopy: string; - -beforeEach(() => { - // gstack-publish reads skills.json relative to the binary's dir (import.meta.dir/..). - // To isolate each test's manifest, we create a sandbox repo that mirrors the real - // structure: copy the bin into sandbox/bin/, write a controlled skills.json at the root. - sandbox = fs.mkdtempSync(path.join(os.tmpdir(), 'publish-sandbox-')); - fs.mkdirSync(path.join(sandbox, 'bin')); - binCopy = path.join(sandbox, 'bin', 'gstack-publish'); - fs.copyFileSync(BIN, binCopy); - fs.chmodSync(binCopy, 0o755); -}); - -afterEach(() => { - fs.rmSync(sandbox, { recursive: true, force: true }); -}); - -function writeManifest(manifest: object): void { - fs.writeFileSync(path.join(sandbox, 'skills.json'), JSON.stringify(manifest, null, 2)); -} - -function writeSkillFile(relPath: string, content = '# Test Skill\n'): void { - const full = path.join(sandbox, relPath); - fs.mkdirSync(path.dirname(full), { recursive: true }); - fs.writeFileSync(full, content); -} - -function run(args: string[]): { status: number | null; stdout: string; stderr: string } { - const result = spawnSync('bun', ['run', binCopy, ...args], { - cwd: sandbox, - encoding: 'utf-8', - timeout: 15000, - }); - return { - status: result.status, - stdout: result.stdout?.toString() ?? '', - stderr: result.stderr?.toString() ?? '', - }; -} - -const VALID_MARKETPLACES = { - fakestore_ok: { - cli: 'true', // binary that always succeeds - login_cmd: 'fakestore_ok login', - publish_cmd_template: 'echo publish {slug} {version}', - docs: 'https://fakestore.example', - auth_check: 'true', // always-authenticated - }, - fakestore_noauth: { - cli: 'true', - login_cmd: 'fakestore_noauth login', - publish_cmd_template: 'echo publish {slug} {version}', - docs: 'https://fakestore.example', - auth_check: 'false', // always-fails auth - }, - fakestore_missing: { - cli: 'nonexistent-binary-xyz', - login_cmd: 'fakestore_missing login', - publish_cmd_template: 'echo publish {slug} {version}', - docs: 'https://fakestore.example', - auth_check: 'nonexistent-binary-xyz whoami', - }, -}; - -function validSkill(slug: string, sourceRel: string, marketplaces: string[] = ['fakestore_ok']) { - const m: Record = {}; - for (const name of marketplaces) m[name] = { slug, publish: true }; - return { - slug, - source: sourceRel, - name: `Skill ${slug}`, - version: '1.0.0', - category: 'test', - description: 'A test skill', - marketplaces: m, - standalone: true, - compatible_hosts: ['claude-code'], - }; -} - -describe('gstack-publish: manifest loading', () => { - test('--list prints every skill and marketplace', () => { - writeSkillFile('skills/alpha/SKILL.md'); - writeSkillFile('skills/beta/SKILL.md'); - writeManifest({ - version: '1.0.0', - description: 't', - skills: [validSkill('alpha', 'skills/alpha/SKILL.md'), validSkill('beta', 'skills/beta/SKILL.md')], - marketplaces: VALID_MARKETPLACES, - }); - const r = run(['--list']); - expect(r.status).toBe(0); - expect(r.stdout).toContain('alpha'); - expect(r.stdout).toContain('beta'); - expect(r.stdout).toContain('fakestore_ok'); - }); - - test('missing manifest exits non-zero', () => { - // Delete any manifest - const r = run(['--dry-run']); - expect(r.status).not.toBe(0); - expect(r.stderr).toContain('skills.json'); - }); - - test('malformed JSON exits non-zero', () => { - fs.writeFileSync(path.join(sandbox, 'skills.json'), '{ not json'); - const r = run(['--dry-run']); - expect(r.status).not.toBe(0); - expect(r.stderr).toContain('parse'); - }); -}); - -describe('gstack-publish: validation', () => { - test('missing source file reports validation error and exits 1', () => { - writeManifest({ - version: '1.0.0', - description: 't', - skills: [validSkill('ghost', 'skills/ghost/DOES_NOT_EXIST.md')], - marketplaces: VALID_MARKETPLACES, - }); - const r = run(['--dry-run']); - expect(r.status).not.toBe(0); - expect(r.stderr).toContain('source file missing'); - expect(r.stderr).toContain('ghost'); - }); - - test('missing slug reports validation error', () => { - writeSkillFile('skills/x/SKILL.md'); - const s = validSkill('temp', 'skills/x/SKILL.md'); - delete (s as Partial).slug; - writeManifest({ - version: '1.0.0', - description: 't', - skills: [s], - marketplaces: VALID_MARKETPLACES, - }); - const r = run(['--dry-run']); - expect(r.status).not.toBe(0); - expect(r.stderr).toContain('missing slug'); - }); - - test('missing version reports validation error', () => { - writeSkillFile('skills/x/SKILL.md'); - const s = validSkill('x', 'skills/x/SKILL.md'); - delete (s as Partial).version; - writeManifest({ - version: '1.0.0', - description: 't', - skills: [s], - marketplaces: VALID_MARKETPLACES, - }); - const r = run(['--dry-run']); - expect(r.status).not.toBe(0); - expect(r.stderr).toContain('missing version'); - }); - - test('no marketplaces configured reports validation error', () => { - writeSkillFile('skills/x/SKILL.md'); - const s = { ...validSkill('x', 'skills/x/SKILL.md'), marketplaces: {} }; - writeManifest({ version: '1.0.0', description: 't', skills: [s], marketplaces: VALID_MARKETPLACES }); - const r = run(['--dry-run']); - expect(r.status).not.toBe(0); - expect(r.stderr).toContain('no marketplaces configured'); - }); -}); - -describe('gstack-publish: dry-run execution', () => { - test('happy path reports DRY-RUN tag and templated command', () => { - writeSkillFile('skills/alpha/SKILL.md'); - writeManifest({ - version: '1.0.0', - description: 't', - skills: [validSkill('alpha', 'skills/alpha/SKILL.md')], - marketplaces: VALID_MARKETPLACES, - }); - const r = run(['--dry-run']); - expect(r.status).toBe(0); - expect(r.stdout).toContain('DRY-RUN'); - expect(r.stdout).toContain('alpha'); - expect(r.stdout).toContain('Published: 1'); - expect(r.stdout).toContain('Failed: 0'); - }); - - test('per-skill filter publishes only the requested slug', () => { - writeSkillFile('skills/alpha/SKILL.md'); - writeSkillFile('skills/beta/SKILL.md'); - writeManifest({ - version: '1.0.0', - description: 't', - skills: [validSkill('alpha', 'skills/alpha/SKILL.md'), validSkill('beta', 'skills/beta/SKILL.md')], - marketplaces: VALID_MARKETPLACES, - }); - const r = run(['alpha', '--dry-run']); - expect(r.status).toBe(0); - expect(r.stdout).toContain('Publishing alpha'); - expect(r.stdout).not.toContain('Publishing beta'); - expect(r.stdout).toContain('Published: 1'); - }); - - test('unknown skill filter exits non-zero', () => { - writeSkillFile('skills/alpha/SKILL.md'); - writeManifest({ - version: '1.0.0', - description: 't', - skills: [validSkill('alpha', 'skills/alpha/SKILL.md')], - marketplaces: VALID_MARKETPLACES, - }); - const r = run(['nonexistent', '--dry-run']); - expect(r.status).not.toBe(0); - expect(r.stderr).toContain('skill not found'); - }); -}); - -describe('gstack-publish: auth check isolation', () => { - test('failing auth for one marketplace does NOT abort the batch in dry-run', () => { - writeSkillFile('skills/alpha/SKILL.md'); - writeManifest({ - version: '1.0.0', - description: 't', - skills: [validSkill('alpha', 'skills/alpha/SKILL.md', ['fakestore_ok', 'fakestore_noauth'])], - marketplaces: VALID_MARKETPLACES, - }); - const r = run(['--dry-run']); - // In dry-run, auth failures are reported but don't block dispatch - expect(r.status).toBe(0); - expect(r.stdout).toContain('fakestore_ok: OK'); - expect(r.stdout).toContain('fakestore_noauth: NOT READY'); - }); - - test('missing binary reported as not-ready with docs link', () => { - writeSkillFile('skills/alpha/SKILL.md'); - writeManifest({ - version: '1.0.0', - description: 't', - skills: [validSkill('alpha', 'skills/alpha/SKILL.md', ['fakestore_missing'])], - marketplaces: VALID_MARKETPLACES, - }); - const r = run(['--dry-run']); - expect(r.stdout).toContain('fakestore_missing: NOT READY'); - expect(r.stdout).toContain('not on PATH'); - }); -}); - -describe('gstack-publish: real manifest sanity', () => { - test('the real repo skills.json passes --dry-run validation', () => { - // This uses the actual bin against the actual manifest (ROOT/skills.json). - // If auth to any real marketplace isn't set up it just reports NOT READY; - // --dry-run still exits 0 because it doesn't require auth to pass. - const real = spawnSync('bun', ['run', path.join(ROOT, 'bin', 'gstack-publish'), '--dry-run'], { - cwd: ROOT, - encoding: 'utf-8', - timeout: 20000, - }); - expect(real.status).toBe(0); - expect(real.stdout).toContain('Validating manifest'); - // Every skill in the real manifest should pass validation - expect(real.stderr).not.toContain('Manifest validation failed'); - }); -});