Tessl • The management layer for AI agent skills

Blog Docs Log inGet started

SkillPublic

harness-ci

Load to wire, tune, or debug a repo's changed-file CI skip.

Quality

61%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./skills/harness-ci/SKILL.md

Quality

Content

57%Weight 40%Scale 1-5

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

The body demonstrates deep domain expertise with concrete commands, flags, and exemplary fail-closed error semantics, but its extreme compression works against it: a 450-word run-on paragraph, repeated links, a scattered main workflow, and two dangling file references (README.md, templates/ci.yml) make it hard to navigate on first read. The material is strong and mostly needs restructuring, not rewriting.

Suggestions

Split the single ~450-word `scripts/change-class` paragraph into headed subsections (verdict classes, render proof, ownership, queue-only behavior) or move the detail into references/wiring.md, leaving a short summary in SKILL.md.

Add an ordered getting-started sequence near the top (wire the workflow once from wiring.md → run the classifier → aggregate required contexts) so the happy path is explicit rather than scattered.

Fix or remove the dangling references to README.md and templates/ci.yml, and consolidate the four repeated links to the same wiring.md section into one pointer.

Dimension	Reasoning	Score
Conciseness	There is no over-explanation of concepts Claude already knows, but the ~450-word single paragraph on `scripts/change-class` is run-on prose, the same wiring.md anchor link is repeated four times, and "Flags and settings: `change-class --help`" deferrals appear twice — clear tightening opportunities. Fits anchor 3 (mostly efficient but could be tightened); not 4 because the repetition and wall-of-text density cost clarity for little token savings.	3 / 5
Actionability	One fully executable command block (`.agents/skills/harness-ci/scripts/harness-only --event pull_request --base "$BASE" --head "$HEAD"`) plus concrete flag documentation (`--mode render-candidate`, `--mode docs`, `--output FILE`, `--lanes-from DIR`, `--queue-only-default true`, `--skippable`, `--waiver`, `--lane JOB=LANE`), with key details legitimately deferred to the bundled scripts' `--help`. Fits anchor 4 (mostly executable guidance with minor gaps); not 5 because the aggregate-needs helper never receives a complete copy-paste example invocation.	4 / 5
Workflow Clarity	Error-path checkpoints are strong (the Fail-closed section, refusal causes carrying remedies such as `cause=source-mirror-missing verb=source operation=refresh`, exit-code semantics, and the Reading a verdict section), but the happy-path sequence (wire the workflow, classify the diff, aggregate required contexts) is scattered across topic sections with no ordered sequence a newcomer can follow. Fits anchor 3 (sequence present but implicit/checkpoints scattered); not 4 because the main workflow steps are never laid out in order.	3 / 5
Progressive Disclosure	`references/wiring.md` exists, is one level deep, and is well signaled with section anchors — but two referenced paths do not resolve in the bundle ([README.md](README.md) and [templates/ci.yml](templates/ci.yml) are missing), and the change-class mega-paragraph inlines detail that belongs in a reference file. Fits anchor 3; not 4 because dangling references and inlined bulk are more than minor organization gaps.	3 / 5
	Total	13 / 20 Passed

Description

53%Weight 40%Scale 1-5

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

The description is short and names a specific niche, but it reads more like an internal mnemonic than a trigger description: the when-clause is implicit, natural synonyms (GitHub Actions, workflows, required checks) are missing, and the skill's concrete capabilities go unmentioned. It would benefit most from an explicit "Use when..." clause and a fuller capability list.

Suggestions

Add an explicit trigger clause, e.g. "Use when wiring or debugging GitHub Actions workflows that skip CI lanes for harness-only or docs-only diffs."

Include natural synonyms users would say: "CI", "GitHub Actions", "workflows", "required checks", "skip tests".

Name the concrete capabilities (classify a CI diff, validate classifier-authorized skipped jobs, aggregate required contexts) instead of the generic verbs "wire, tune, debug".

Dimension	Reasoning	Score
Specificity	"wire, tune, or debug" are concrete verbs on a named domain ("a repo's changed-file CI skip"), but the description conveys only a thin slice of the skill's actual capabilities (classify diffs, validate skipped jobs, aggregate required contexts), and "tune" is generic. Matches anchor 3 (names domain and 1-2 concrete actions, not comprehensive); not 4 because coverage of the action set is minimal.	3 / 5
Completeness	The "what" is clear ("wire, tune, or debug a repo's changed-file CI skip"), but the "when" is only weakly implied by the "Load to..." phrasing with no explicit "Use when..." trigger clause, which caps completeness at 3 per the judging guidelines. Not 4 because the when-guidance is not explicit.	3 / 5
Trigger Term Quality	"CI", "skip", and "repo" are terms a user would naturally say, but common variations and synonyms are absent: "GitHub Actions", "workflows", "required checks", "skip tests". Fits anchor 3 (some relevant keywords, missing common variations); not 4 because keyword coverage has clear gaps.	3 / 5
Distinctiveness Conflict Risk	"changed-file CI skip" is a narrow niche phrase unlikely to trigger unrelated skills. Fits anchor 4 (mostly distinct, minor overlap risk with closely related CI-automation skills); not 5 because it does not enumerate distinct trigger phrases.	4 / 5
	Total	13 / 20 Passed

Validation

75%

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 — 12 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 2 missing, 4 suspicious	Warning
referenced_paths_exist	Referenced path issues: 2 deeper-than-1-level	Warning

	Total	12 / 16 Passed

Repository: vanillagreencom/kendex
Path: skills/harness-ci/SKILL.md
Revision: 32d7bc7

Reviewed: 13 hours ago

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.

harness-ci

Table of Contents

Is this your skill?