ts-sdk-author

Design, build, verify, and publish production-grade TypeScript SDKs as npm packages inside a pnpm monorepo. Covers workspace layout, public API and module boundaries, plugin extension points, branded types and library-tuned tsconfig, tsdown bundling (vs tsup/tsc-only/unbuild), package.json exports with dual ESM+CJS and isomorphic conditions (browser/workers/RN/deno), Turborepo pipelines, publint and @arethetypeswrong/cli verification, changesets pre-release mode, npm dist-tags (latest/next/beta/rc/canary), and the alpha→beta→rc→stable release lifecycle. Triggers on: build a TS SDK, extract core library, package.json exports, dual ESM CJS, tsdown config, tsup vs tsdown, publint, attw, changesets prerelease, npm dist-tag, beta to rc, canary release, pnpm workspace SDK, isomorphic SDK, tsconfig library, npm provenance, shipping a TypeScript library.

Quality

92%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

85%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A highly actionable, well-sequenced SDK-authoring workflow with strong progressive disclosure into seven real reference files and explicit verification checkpoints. Its only real weakness is conciseness: a provenance "Source Skills" section and some restating prose add tokens that do not aid execution.

Suggestions

Remove or compress the trailing "Source Skills" section — the composed-from/sources provenance does not help Claude execute the workflow and is already captured in frontmatter metadata.

Trim the opening paragraph ("End-to-end workflow for shipping...") and lines like "The seven references hold the depth" since they restate the description; let the phased workflow stand on its own.

Cut evaluative filler such as "This is where most TS SDK bugs live" / "This is where most SDK projects accumulate debt" in favor of letting the invariants and checklists speak for themselves.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient and code/config-dense, but it carries content Claude does not need — the closing "Source Skills" provenance section, an intro paragraph that restates the description, and some commentary ("This is where most TS SDK bugs live") — so it sits at "mostly efficient but could be tightened" rather than "every token earns its place". Not a 1 because it does not explain basic concepts Claude already knows.	2 / 3
Actionability	Packed with copy-paste-ready, executable artifacts — a full tsconfig.build.json, tsdown.config.ts, package.json exports blocks, turbo.json, and bash commands for changesets/dist-tags/publishing — matching the anchor for fully executable code with specific examples.	3 / 3
Workflow Clarity	A clearly sequenced seven-phase workflow with an explicit verification phase (publint/attw/smoke tests wired into prepublishOnly), failure-mode explanation ("Masquerading ESM"), and a comprehensive pre-publish checklist, satisfying the anchor for clear sequence with explicit validation steps and checklists.	3 / 3
Progressive Disclosure	SKILL.md is an overview that points to seven real, one-level-deep reference files (all present in ./references/), each signaled with a "Read next:" pointer plus specific section numbers (e.g. §2, §3) and summarized in a Reference Files table — textbook clear-overview with well-signaled references and easy navigation.	3 / 3
	Total	11 / 12 Passed

Description

100%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, third-person description that states concrete capabilities and pairs them with an explicit trigger list, cleanly answering both what the skill does and when to invoke it. Voice is correct (no first/second person) and the niche is highly distinctive.

Dimension	Reasoning	Score
Specificity	Lists many concrete actions — "Design, build, verify, and publish production-grade TypeScript SDKs" plus a detailed coverage list (workspace layout, module boundaries, branded types, tsdown bundling, dual ESM+CJS exports, Turborepo pipelines, publint/attw, changesets, dist-tags, release lifecycle), matching the anchor for multiple specific concrete actions.	3 / 3
Completeness	Clearly answers both what (the lead verb chain plus the "Covers" enumeration) and when (an explicit "Triggers on:" clause), satisfying the top anchor for both what AND when.	3 / 3
Trigger Term Quality	The explicit "Triggers on:" list surfaces natural developer phrasings — "build a TS SDK", "tsup vs tsdown", "publint", "attw", "beta to rc", "canary release", "pnpm workspace SDK" — giving good coverage of terms users would actually say.	3 / 3
Distinctiveness Conflict Risk	The niche is sharply defined — production-grade TS SDK authoring in a pnpm monorepo with a specific toolchain (tsdown/changesets/attw) — and the triggers are distinctive enough that it is unlikely to fire for unrelated skills.	3 / 3
	Total	12 / 12 Passed

Validation

87%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (603 lines); consider splitting into references/ and linking	Warning
metadata_field	'metadata' should map string keys to string values	Warning

	Total	14 / 16 Passed

Repository: mindfold-ai/Trellis
Path: .agents/skills/ts-sdk-author/SKILL.md
Commit: 12e279a

Reviewed: 5 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.