CtrlK
BlogDocsLog inGet started
Tessl Logo

cache-deception

Web cache deception — trick CDN/proxy into caching authenticated responses under unauthenticated URLs, exposing PII to any visitor.

60

Quality

70%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Critical

Do not install without reviewing

Fix and improve this skill with Tessl

tessl review fix ./packages/decepticon/decepticon/skills/standard/exploit/web/cache-deception/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

86%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 tight, highly actionable single-file skill with concrete probe commands, an executable PoC, a validation checkpoint, and a defender section. Minor intro restatement is the only conciseness drag, and error-recovery guidance is light.

Suggestions

Trim the intro paragraph so it does not restate the frontmatter description, and drop filler like the "Each cache impl handles these differently... known quirks" line.

Add a short "if step 2 does NOT return authenticated content" branch (try alternate suffixes from the variants table, or conclude not vulnerable) to strengthen the workflow feedback loop.

Clarify that the Python PoC asserts are illustrative markers to adapt per target rather than literal success gates, so the workflow checkpoint is unambiguous.

DimensionReasoningScore

Conciseness

The body is dense and mostly lean — variant/severity tables and code blocks earn their tokens — with only minor restatement (the intro re-explains the description) and a couple of filler sentences that could be trimmed; anchor 5 requires every token to earn its place, so 4 fits better than 5.

4 / 5

Actionability

It provides copy-paste-ready curl probe commands, a complete executable Python PoC with asserts, a concrete variant-path table, and an nginx defender config covering the common cases.

5 / 5

Workflow Clarity

The probe section sequences Step 1 (authenticated) → Step 2 (unauthenticated) → "If the second request returns the AUTHENTICATED content of step 1 → deception confirmed" as an explicit validation checkpoint, but error-recovery guidance when the check fails is limited; anchor 5 needs feedback loops, so 4 fits better than 5.

4 / 5

Progressive Disclosure

No bundle files exist, so the skill is a single well-organized file with clearly headed sections (## 1–## 8) and a Cross-references block pointing one level deep to the upstream catalog and smuggling.md, making navigation easy.

5 / 5

Total

18

/

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 gives a clear, specific mechanism and outcome for web cache deception but omits any explicit "Use when..." trigger guidance, which caps completeness and limits trigger coverage. Vendor and synonym terms exist only in metadata rather than the description itself.

Suggestions

Add an explicit trigger clause, e.g. "Use when testing CDN/proxy caching for authenticated-content exposure or when the user mentions cache deception, Cloudflare/CloudFront/Akamai, or path-normalization caching bugs."

Fold 1-2 high-signal vendor keywords (cloudflare, cloudfront, akamai) from metadata.when_to_use into the description so natural trigger terms are present there.

Briefly distinguish from cache poisoning in the description ("deception caches victim data; poisoning injects content") to reduce overlap with the related poisoning class.

DimensionReasoningScore

Specificity

The description names the domain and one concrete action — "trick CDN/proxy into caching authenticated responses under unauthenticated URLs" — with the outcome "exposing PII", but it is a single technique rather than a list of multiple specific actions.

3 / 5

Completeness

It clearly states what the skill does, but there is no "Use when..." clause or equivalent trigger guidance in the description field, which caps completeness at 3 per the rubric guideline.

3 / 5

Trigger Term Quality

It includes the natural term "Web cache deception" plus "CDN/proxy" and "PII", but vendor synonyms (cloudflare, cloudfront, akamai) and "cache poisoning" live in metadata.when_to_use rather than the description, leaving common variations missing.

3 / 5

Distinctiveness Conflict Risk

"Web cache deception" is a clear, well-known niche with minimal conflict risk, though it sits adjacent to cache poisoning and the description does not explicitly distinguish the two here.

4 / 5

Total

13

/

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.

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

Total

15

/

16

Passed

Repository
PurpleAILAB/Decepticon
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.