sensei

Iteratively improve skill frontmatter to achieve good routing test coverage. WHEN: run sensei, sensei help, improve skill routing

Quality

75%

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 ./.github/skills/sensei/SKILL.md

Quality

Content

77%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 is a well-structured, actionable instruction skill with an explicit validate-and-repeat workflow and proper one-level-deep reference usage. Its main weakness is token redundancy: the ASCII help box duplicates the Main Loop steps it precedes.

Suggestions

Slim the Help box to usage/examples only and drop the 8-step WHAT IT DOES list, since the identical steps appear in the Main Loop.

Replace vague sub-steps like "If there are failed tests, suggest fixes" with the specific fix procedure or a pointer to it.

Clarify what "unimplemented suggestions" in the SUMMARY step means, e.g. convention violations the user declined to auto-fix.

Dimension	Reasoning	Score
Conciseness	The 24-line ASCII help box restates the same 8 steps that the Main Loop section immediately re-describes in detail — a noticeable duplication that could be tightened to a short usage summary. Otherwise the body is efficient with no filler, so it is above anchor 2 but not the 'minor trim' level of anchor 4.	3 / 5
Actionability	Guidance is concrete for an instruction-only skill: exact paths ("plugins/{plugin}/skills/{skill-name}/SKILL.md", "evals/{skill-name}/*.yaml"), a linked conventions file, and a bounded iteration count. Minor gaps like "If there are failed tests, suggest fixes" leave it just below fully executable anchor 5.	4 / 5
Workflow Clarity	The 8-step Main Loop is clearly sequenced with an explicit validation step (TEST), a feedback loop (REPEAT to step 2 "until the frontmatter aligns... AND routing tests pass"), a hard bound (max 5 iterations), and a user checkpoint (PROMPT) — matching the anchor-5 pattern of validation plus error-recovery loops.	5 / 5
Progressive Disclosure	Sections (Help, Main Loop, Constraints, Related Skills) are well organized, and the one detailed reference (references/CONVENTIONS.md, verified to exist) is clearly signaled with a markdown link one level deep; the vally-eval dependency is surfaced in Related Skills rather than inlined.	5 / 5
	Total	17 / 20 Passed

Description

62%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 answers both what and when with an explicit WHEN clause and a distinctive name-based trigger, but the capability list is thin and the triggers lean heavily on the skill's own name rather than natural user phrases. It sits comfortably above the vague-examples band but below the exemplar descriptions.

Suggestions

Expand the 'what' to enumerate the concrete pipeline actions, e.g. "Verifies frontmatter against conventions, scaffolds routing tests, and iterates until they pass".

Add natural trigger phrases users would actually say beyond the skill name, such as "fix skill frontmatter", "add WHEN triggers", or "routing tests are failing".

Include a clarifying second sentence (per the CONVENTIONS.md format) naming the target artifact, e.g. what counts as routing test coverage.

Dimension	Reasoning	Score
Specificity	"Iteratively improve skill frontmatter" and "achieve good routing test coverage" name the domain plus 1-2 concrete actions, but the description omits the verify/scaffold/test steps the skill actually performs, so coverage is not comprehensive as required for anchor 4.	3 / 5
Completeness	Both a clear "what" ("Iteratively improve skill frontmatter to achieve good routing test coverage") and an explicit WHEN clause are present, but the WHEN is narrow and name-dependent rather than the comprehensive concrete trigger phrases of anchor 5.	4 / 5
Trigger Term Quality	The WHEN triggers ("run sensei, sensei help, improve skill routing") are relevant but two of three depend on the skill's own name, and common natural variations users would say ("fix skill frontmatter", "add WHEN triggers", "skill routing tests") are missing.	3 / 5
Distinctiveness Conflict Risk	"sensei" and "improve skill routing" carve out a distinct niche with minimal conflict risk, though "improve skill frontmatter" could mildly overlap with generic skill-maintenance skills.	4 / 5
	Total	14 / 20 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: microsoft/GitHub-Copilot-for-Azure
Path: .github/skills/sensei/SKILL.md
Commit: fcf2f3b

Reviewed: 6 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.