CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

avoid-ai-writing-router

Use when a request combines AI-writing audit, rewrite, file editing, voice preservation, false-positive interpretation, verification, or when the user invokes Avoid AI Writing without naming a mode.

Quality

76%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./skills/avoid-ai-writing-router/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

81%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 router body is a strong example of an instruction-only orchestration skill: lean operational rules, unambiguous routing, typed edges, and a verification-gated workflow with a bounded repair loop. Its only notable weaknesses are a missing example/template for the handoff envelope and two bundle scripts that lack any navigation path from the skill document.

DimensionReasoningScore

Conciseness

The body is dense, rule-only prose with no explanations of concepts Claude already knows ("Classify the request once, create the smallest useful handoff envelope", "Do not cycle indefinitely"). Minor redundancy: the editing-pass budget rules are restated across 'Multi-stage sequencing', 'Boundary changes', and 'Stop conditions', which keeps it just below anchor 5.

4 / 5

Actionability

Concrete, executable routing guidance throughout: numbered rules mapping request types to named skills ("Returned-text rewrite... go to `voice-preserving-rewriter`"), typed edges with exact semantics, exact field names (`human_representation_sensitive`, `protected_constraints`), and a worked multi-stage example ("scan this, rewrite it, and make sure nothing important changed"). The one gap: the handoff envelope — the skill's central artifact — is defined only as a bulleted field list with no template or example instance, so anchor 5's 'specific examples cover the common cases' is not fully met.

4 / 5

Workflow Clarity

The multi-stage sequence has explicit validation checkpoints and a bounded feedback loop: "preservation-verifier checks before/after material", "A verifier FAIL returns to the correct repair owner once only if the shared budget has room", re-verification after repair, explicit stop conditions, and honest failure reporting ("Report an exhausted-budget failure without calling the result verified"). The destructive-operation cap does not apply because file mutation is guarded by user authorization plus preservation verification.

5 / 5

Progressive Disclosure

The body is a ~120-line overview with well-signaled, one-level-deep references, each purpose-labeled in the authority chain ("references/handoff-contract.md defines what context and evidence can cross Skill boundaries"); all four referenced files exist and skill-graph.json is valid JSON containing the claimed nodes, edges, guards, and loop limits. The gap: scripts/validate_connections.py and its test exist in the bundle but are referenced neither in SKILL.md nor in any reference file, leaving two bundle files undiscoverable — a minor organization gap that keeps it below anchor 5.

4 / 5

Total

17

/

20

Passed

Description

61%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 has strong, explicit trigger guidance with a clear niche, but it is entirely 'when' and no 'what': it never states that the skill routes and coordinates the plugin's specialized stages. Adding an explicit capability statement and more natural user phrasings like 'humanize' would lift specificity and completeness.

Suggestions

Lead with an explicit capability statement (e.g., 'Routes and coordinates the Avoid AI Writing plugin's specialized skills for detection, rewriting, and verification.') before the 'Use when' clause so the 'what' is stated, not implied.

Add natural trigger phrasings users actually say, such as 'humanize', 'make this sound human / less AI-generated', or 'was this written by AI'.

Replace internal jargon like 'false-positive interpretation' with user-facing wording (e.g., 'claims that text was AI-written') so the trigger matches how users phrase consequential authorship questions.

DimensionReasoningScore

Specificity

The description lists six concrete activity areas ("AI-writing audit, rewrite, file editing, voice preservation, false-positive interpretation, verification"), but these are the user's request types, not the skill's own actions — the actual capability (routing/orchestration across the plugin's skills) is never stated. It exceeds 'names domain and 1-2 concrete actions' but falls short of anchor 4 because no capability verb like 'routes' or 'coordinates' appears.

3 / 5

Completeness

The 'when' is explicit and specific ("Use when a request combines... or when the user invokes Avoid AI Writing without naming a mode"), but the 'what' is only implicit — the description never states what the skill does (coordinate/route the specialized skills). Better than anchor 2 because the enumerated activities convey the domain beyond a bare 'Use when working with documents', but below anchor 4 because the skill's function is absent.

3 / 5

Trigger Term Quality

Good keyword coverage across the domain's request verbs ("AI-writing audit", "rewrite", "file editing", "verification") plus an explicit name-invocation trigger ("when the user invokes Avoid AI Writing without naming a mode"). Missing a few natural user phrasings such as 'humanize', 'sounds AI-generated', or 'make it sound human', which keeps it below anchor 5.

4 / 5

Distinctiveness Conflict Risk

Specialized vocabulary ("voice preservation", "false-positive interpretation") and the mode-disambiguation clause ("without naming a mode") carve a clear niche and reduce sibling conflicts. Not anchor 5 because the listed activities are exactly the triggers of the sibling skills it routes to (ai-writing-detector, voice-preserving-rewriter), leaving minor overlap risk with closely related skills.

4 / 5

Total

14

/

20

Passed

Validation

93%

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

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository

conorbronsdon/avoid-ai-writing

Path

skills/avoid-ai-writing-router/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.