CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

cloudflare

Comprehensive Cloudflare platform skill covering Workers, Pages, storage (KV, D1, R2), AI (Workers AI, Vectorize, Agents SDK), networking (Tunnel, Spectrum), security (WAF, DDoS), and infrastructure-as-code (Terraform, Pulumi). Use for any Cloudflare development task.

53

Quality

67%

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 ./packages/opencode/test/fixture/skills/cloudflare/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

46%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-organized router with genuinely useful decision trees, but it is duplicated between the trees and the product index, and — critically — it is an index to a bundle that is not present: every referenced path is dead and the frontmatter's 5-entry reference list contradicts the ~60 references the body claims. As written, the skill's core promise ('load detailed references') cannot be fulfilled.

Suggestions

Ship the actual reference files (or trim the index to only the references that exist) — the entire body routes to `references/<product>/` paths that are absent from the bundle.

Reconcile the frontmatter `references` list (5 entries) with the body's ~60 referenced products, so the declared bundle matches what the body navigates to.

Remove the duplication between the Quick Decision Trees and the Product Index tables — one mapping (the tables, annotated with the decision-tree hints) would carry the same information in half the tokens.

DimensionReasoningScore

Conciseness

There is no explanation of concepts Claude already knows — good — but the same routing information is stated twice: the decision trees ('├─ Relational SQL → d1/ (SQLite) or hyperdrive/') and then the Product Index tables repeat the identical product→reference mapping, and the frontmatter description repeats the category taxonomy a third time. This matches 'mostly efficient but ... could be tightened'; it is not a 2 because nothing is padded prose, and not a 4 because the duplication roughly doubles the body's length.

3 / 5

Actionability

The guidance is concrete as routing ('Strongly-consistent per-entity state → durable-objects/ (DO storage)'), but the one action the skill instructs — 'then load detailed references' — cannot be executed: no references/ directory exists, so every path like `references/workers/` is dead. This fits 'some concrete guidance but incomplete / missing key details'; not a 2 because the decision trees themselves give specific, usable product selection, not just high-level hints.

3 / 5

Workflow Clarity

The intro states a clear two-step sequence — 'Use decision trees below to find the right product, then load detailed references' — which is unambiguous routing. It sits at 3 rather than 4 because the second step is unexecutable (the referenced files do not exist) and there is no guidance on what to do once a reference is loaded, leaving the sequence with a gap at its terminal step.

3 / 5

Progressive Disclosure

The internal structure is well designed — clear sections, one-level-deep references, easy navigation — but scoring against the actual bundle, none of the ~60 referenced paths (e.g. `references/workers/`, `references/vectorize/`) exist on disk, and the frontmatter lists only 5 references (workers, pages, d1, durable-objects, workers-ai) contradicting the body's full index. Navigation that resolves to nothing fits 'minimal structure / references buried' territory; it is above 1 only because the routing layout itself is genuinely clear.

2 / 5

Total

11

/

20

Passed

Description

75%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 solid description with explicit what/when structure and dense, natural product keywords. Its weaknesses are the buzzword opener 'Comprehensive', a 'when' clause that over-claims scope ('any Cloudflare development task'), and omission of the most common colloquial trigger, 'wrangler'.

Suggestions

Drop 'Comprehensive' and lead with the concrete coverage: 'Build and deploy on Cloudflare: Workers, Pages, KV, D1, R2, Workers AI, Vectorize, WAF, Terraform/Pulumi.'

Add high-frequency trigger terms like 'wrangler', 'Durable Objects', and 'edge deployment' to the description so it matches how users actually phrase Cloudflare requests.

Narrow the 'when' clause from 'any Cloudflare development task' to concrete triggers such as 'Use when the user mentions Cloudflare, wrangler, Workers, Pages, or deploying to the edge.'

DimensionReasoningScore

Specificity

The description enumerates many concrete capabilities — 'covering Workers, Pages, storage (KV, D1, R2), AI (Workers AI, Vectorize, Agents SDK), networking (Tunnel, Spectrum), security (WAF, DDoS), and infrastructure-as-code (Terraform, Pulumi)' — giving concrete coverage across categories. It falls below a 5 because it lists product names rather than actions ('covering' is not a verb like 'extract' or 'deploy'), and 'Comprehensive' is mild over-claim padding.

4 / 5

Completeness

Both 'what' (the enumerated product coverage) and 'when' ('Use for any Cloudflare development task') are explicitly present, matching the score-4 anchor. It is not a 5 because the 'when' clause is maximally broad ('any ... development task') rather than a concrete trigger phrase, and not a 3 because the trigger guidance is explicit, not merely implied.

4 / 5

Trigger Term Quality

Strong natural keyword coverage: 'Cloudflare', 'Workers', 'Pages', 'KV', 'D1', 'R2', 'Workers AI', 'Vectorize', 'Tunnel', 'WAF', 'Terraform', 'Pulumi' are terms users would actually say. Not a 5 because common variations are missing — notably 'wrangler' (the CLI nearly every Cloudflare task mentions), 'Durable Objects', and any file/deployment artifacts like '.workers.dev'.

4 / 5

Distinctiveness Conflict Risk

'Cloudflare' carves out a clear niche with distinct product-specific triggers, so conflict risk is low — matching 'mostly distinct; minor overlap risk'. Minor residual overlap: Terraform/Pulumi are whole ecosystems that could have their own dedicated skills and would also match this description.

4 / 5

Total

16

/

20

Passed

Validation

87%

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

Validation — 14 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

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

Warning

referenced_paths_exist

Referenced path issues: 60 missing

Warning

Total

14

/

16

Passed

Repository

elastic/elastic-ramen

Path

packages/opencode/test/fixture/skills/cloudflare/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.