sdk-seam

Discipline for the seam between two SDKs (or two sides of one contract) that the same hand writes. The failure mode: "we own both sides" produces dirty contracts no foreign reviewer would accept. The exercise: pretend the other side is FFI, IPC, or a network protocol you cannot rewrite. Spawn an adversarial subagent profiled as the producer's maintainer; negotiate the change as a feature request, not a PR. Companion to $sdk-design. Language-agnostic — applies to a TS package + its consumer, a Rust crate + its WASM binding, two services sharing a wire format, or any other boundary the same author writes both ends of.

Quality

72%

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

Fix and improve this skill with Tessl

tessl review fix ./.agents/skills/sdk-seam/SKILL.md

Quality

Content

77%

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

Highly actionable and clearly sequenced with strong feedback loops, but somewhat verbose through restatement and monolithic — the long body and reusable templates could be split into one-level-deep reference files.

Suggestions

Tighten or cut the redundant restatement sections ('The short version', 'Why this is more than ceremony') to reduce tokens without losing guidance.

Extract the reusable subagent brief and the FEEDBACKS.md MUST/MUST NOT template into a reference file one level deep, keeping SKILL.md as an overview that links to them.

Dimension	Reasoning	Score
Conciseness	Mostly efficient prose that doesn't re-explain basics like FFI/IPC, but it pads with rhetorical emphasis and restatement — 'The short version' and 'Why this is more than ceremony' restate earlier sections, and 'That's not a virtue. It's a warning.'-style lines add tokens without new guidance.	2 / 3
Actionability	Provides copy-paste-ready, concrete guidance for an instruction-only skill: a literal subagent prompt template, a MUST/MUST NOT FEEDBACKS.md structure, a four-item restate exercise with a worked example, and four procedural stages.	3 / 3
Workflow Clarity	Clear multi-step sequence (flow diagram + Steps 1–3, contract-first staging) with explicit validation checkpoints ('Producer tests run; consumer tests run', 'Lock the contract... then update consumers') and a refuse/rewrite/re-spawn feedback loop.	3 / 3
Progressive Disclosure	Well-organized with clear headers, but it is a single ~330-line monolithic file with no bundle files; content like the subagent brief or FEEDBACKS.md template that could live one level deep in a reference is inline, and only external skill companions ($sdk-design, $etiology, $pedantic) are referenced.	2 / 3
	Total	10 / 12 Passed

Description

67%

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 specific, distinctive description with concrete actions, but it omits an explicit 'Use when' trigger clause and leans on technical jargon over natural user keywords.

Suggestions

Add an explicit 'Use when...' trigger clause naming the natural situations a user would describe (e.g., when changing both ends of an SDK boundary, a crate and its WASM binding, or a shared wire format).

Soften the jargon density with one or two phrases a user would actually say ('keeping a two-sided API clean', 'same author on both sides of a contract').

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'Spawn an adversarial subagent profiled as the producer's maintainer; negotiate the change as a feature request, not a PR' and 'pretend the other side is FFI, IPC, or a network protocol you cannot rewrite' — not vague abstractions.	3 / 3
Completeness	It clearly states what the skill does (discipline for keeping same-author seams clean) but lacks an explicit 'Use when...' clause; the 'when' is only implied via the applies-to enumeration, which caps completeness at 2 per the rubric.	2 / 3
Trigger Term Quality	Relevant boundary terms appear (FFI, IPC, WASM binding, Rust crate, wire format) but they lean technical-jargon rather than natural user phrasing, and common variations of how a user would actually request this are thin.	2 / 3
Distinctiveness Conflict Risk	A clear narrow niche — 'the seam between two SDKs (or two sides of one contract) that the same hand writes' with the adversarial-subagent move — unlikely to trigger for an unrelated skill.	3 / 3
	Total	10 / 12 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
relative_links	Relative link issues: 1 suspicious	Warning

	Total	15 / 16 Passed

Repository: gridaco/grida
Path: .agents/skills/sdk-seam/SKILL.md
Commit: 2e0d276

Reviewed: 3 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.