CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

issue-creation

Trigger: issue creation, bug reports, feature requests, or issue approval. Create and triage GitHub issues from repository evidence.

Quality

81%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

SKILL.md
Quality
Evals
Security

Quality

Content

73%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 delivers an exceptionally rigorous, validation-heavy workflow with executable gh CLI commands, precise read-back verification, and fail-closed error handling. Its weaknesses are density and redundancy — rules are restated across sections, sentences are over-stuffed, and repository-specific policy is inlined rather than offloaded to the reference file.

Suggestions

Deduplicate rules repeated across Hard Rules, Approved Catalog Gate, and Execution Steps (duplicate search, label permissions, protected-label handling) into one canonical section and cross-reference it.

Replace the step-6 label-expansion table with a one-line rule (append one `--label "$PERMITTED_LABEL"` pair per permitted label, none if zero) and split long semicolon-chained sentences into short bullets.

Move the repository-specific gentle-ai catalog policy (the two paragraphs naming github.com/Gentleman-Programming/gentle-ai) into a reference file like references/delegated-workflow-actions.md to slim the generic body.

DimensionReasoningScore

Conciseness

No tokens are spent explaining concepts Claude already knows, but rules are restated across sections (duplicate search in Hard Rules and Execution Steps, protected-label and capability rules in both Hard Rules and Approved Catalog Gate), the step-6 table overspecifies trivial `--label` token expansion, and long semicolon-chained sentences could be tightened. It is not 4 because the redundancy exceeds minor over-explanation; it is not 2 because there is no padding or filler — every sentence carries operative rule content.

3 / 5

Actionability

Concrete executable `gh` commands with exact flags cover the main operations (label enumeration via `gh api`, `gh issue list --state all`, `gh issue view --json`, `gh issue create --body-file`, comment read-back with identity verification). It is not 5 because derivation of key inputs like `$QUERY`, `$TITLE`, and the "reviewed answers" is left abstract and the conforming-equivalent comment path never shows its command; it is not 3 because the shown commands are real, executable, and span the critical mutation and verification cases.

4 / 5

Workflow Clarity

Nine explicitly sequenced steps with validation checkpoints at every phase: target identity verified from `DISCOVERY_FILE` and `READBACK_FILE`, exact title/body comparison after normalization, and a fail-closed `confirmed | no_write | unknown` classification where `unknown` halts all later mutations and retries. It is not 4 because no checkpoint gaps are evident — this is a write-operation skill whose validation is thorough, so the missing-validation cap does not apply.

5 / 5

Progressive Disclosure

The single bundle file `references/delegated-workflow-actions.md` exists, is clearly signaled as normative ("read ... completely and follow it. It is normative, not optional background"), and is one level deep (87 lines, no nested references). It is not 5 because the main body inlines a large mass of dense policy detail (Hard Rules, Approved Catalog Gate) that could be split out, leaving the overview heavier than an overview should be; it is not 3 because sections are well-organized and the reference path is explicit rather than buried.

4 / 5

Total

16

/

20

Passed

Description

78%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.

A compact, third-person description that answers both what and when with concrete trigger phrases and a well-defined GitHub-issues niche. Its main limitation is that the 'what' clause captures only two of the skill's actual capabilities, and a few natural trigger variations (open an issue, file a bug, ticket) are absent.

DimensionReasoningScore

Specificity

The 'what' is "Create and triage GitHub issues from repository evidence" — the GitHub-issues domain plus two concrete actions (create, triage). It is not 4 because several real capabilities (commenting, labeling, duplicate handling, workflow approvals) are absent from the description; it is not 2 because the actions named are concrete rather than generic. Voice is third person, so no penalty applies.

3 / 5

Completeness

Both parts are explicit: the what ("Create and triage GitHub issues from repository evidence") and the when ("Trigger: issue creation, bug reports, feature requests, or issue approval") with concrete trigger phrases. Equivalent explicit trigger guidance exists, so the missing-'Use when' cap does not apply; it is not 4 because the when clause is specific and concrete rather than weakly implied.

5 / 5

Trigger Term Quality

Natural phrases a user would say are present: "issue creation", "bug reports", "feature requests", "issue approval", "GitHub issues". It is not 5 because common variations like "open an issue", "file a bug", or "ticket" are missing; it is not 3 because coverage goes beyond one or two keywords and spans both task nouns and the domain noun.

4 / 5

Distinctiveness Conflict Risk

The description carves a clear GitHub-issues niche anchored by "GitHub issues", keeping conflict risk low. It is not 5 because the bare trigger phrases "bug reports" and "feature requests" could plausibly fire a generic bug-writing or report-drafting skill before the GitHub anchor applies; it is not 3 because the niche is otherwise well-differentiated.

4 / 5

Total

16

/

20

Passed

Validation

100%

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

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

Gentleman-Programming/gentle-ai

Path

internal/assets/skills/issue-creation/SKILL.md

Reviewed

Table of Contents

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.