CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

web-subdomain-takeover

Subdomain takeover via dangling DNS/CNAME — GitHub Pages, Heroku, Azure, Fastly, Shopify, Netlify, Surge, Tumblr, Beanstalk, Zendesk, etc.

Quality

77%

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/recon/web-recon/subdomain-takeover/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.

A strong, dense operational skill: nearly every section is executable, the workflow is explicitly sequenced with a decision gate and manual hit verification, and the non-obvious knowledge (fingerprints, tool invocations, OPSEC boundaries for the PoC) is exactly what a skill should carry. The main improvements are trimming the known-concept explanations in the intro and chains table, showing one concrete provider claim example, and moving the provider fingerprint table to a references/ file.

Suggestions

Move the 25-provider fingerprint table to references/providers.md and reference it from sections 2 and 3, keeping SKILL.md as a lean overview — this would also let the section-3 grep list be generated from one authoritative source instead of a hardcoded subset.

Show one worked claim example (e.g., GitHub Pages: create repo named <user>.github.io, enable Pages, push the marker page) so the payoff step is as concrete as the detection steps, instead of "Deploy via the provider's normal flow".

Tighten the intro paragraph and Section 6 chains table to severity-justifying bullets — Claude already knows cookie scoping, OAuth redirect_uri, and CSP semantics — and cut or compress Section 7's defender signatures if the skill is scoped to offensive/bug-bounty use.

DimensionReasoningScore

Conciseness

The body is dominated by executable commands and genuinely non-obvious data (fingerprint strings like "There isn't a GitHub Pages site here", "The gods are wise, but do not know of the site which you seek", and tool flags), but a few passages re-explain what Claude already knows: the intro paragraph spelling out cookie/CORS/OAuth/CSP inheritance, and the Section 6 chains table walking through cookie theft, OAuth redirect_uri, and CSP bypass mechanics. This fits anchor 4 ('efficient; minor instances of over-explanation that could be trimmed') rather than level 5, while being far from level 2-3 verbosity.

4 / 5

Actionability

Nearly everything is copy-paste ready: subfinder/amass/dnsx enumeration, a dig-based CNAME extraction loop, curl+grep fingerprint detection, subjack/subzy/nuclei mass-scan invocations, and a heredoc marker page with a proof curl. The one minor gap is the claim step — "Deploy via the provider's normal flow (git push, heroku create, etc.)" defers the actual claim to the upstream reference rather than showing a concrete example for even one common provider (e.g., GitHub Pages). This matches anchor 4 ('mostly executable guidance; concrete code or commands with minor gaps'), just short of level 5's fully covered common cases.

4 / 5

Workflow Clarity

The skill has a clearly numbered pipeline (enumerate → fingerprint → detect → mass-detect → claim → report) with explicit validation and feedback loops: a "Manual sanity check on hits" loop after mass scanning, NXDOMAIN verification at the DNS layer for providers with no body fingerprint, false-positive handling ("Often false positive — verify provider docs before claiming"), and a full Section 8 decision-gate checklist mapping observations to actions including a stop condition when a program forbids takeover testing. This matches anchor 5 ('clear sequence with explicit validation steps; feedback loops for error recovery; checklists for complex processes') — the batch-scan validation guideline is satisfied, so no cap applies.

5 / 5

Progressive Disclosure

The single file is well-organized with numbered, clearly headed sections and only one-level-deep, clearly signaled cross-references (sister skills and the upstream can-i-take-over-xyz repo). However, at ~165 lines with a 25-provider fingerprint table inlined, some content that would fit a references/ file (the provider table, which is reference data rather than workflow instruction) is inline, and no bundle split exists at all. This fits anchor 4 ('good structure; most content appropriately placed; minor organization gaps') rather than the under-50-line simple-skill exception to level 5.

4 / 5

Total

17

/

20

Passed

Description

65%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 distinctive, domain-specific description with strong natural trigger phrases and a clear 'what', but it lacks an explicit 'when to use' clause, and the excellent trigger vocabulary in metadata.when_to_use (subjack, subzy, can-i-take-over-xyz, orphaned subdomain) is not surfaced in the description field itself. Adding a 'Use when...' sentence would lift both completeness and trigger coverage.

Suggestions

Append an explicit trigger clause to the description, e.g.: "Use when auditing subdomains for dangling CNAME/DNS takeover claims, or when scans flag NXDOMAIN provider targets, orphaned subdomains, or takeover hits from subjack/subzy/nuclei."

Fold the natural trigger terms currently buried in metadata.when_to_use ("dangling CNAME", "orphaned subdomain", "subjack", "subzy", "can-i-take-over-xyz") into the description text itself so they surface at trigger time.

State 1-2 concrete actions (e.g., "Enumerate subdomains, fingerprint dangling CNAMEs, and verify claimability across 25+ providers") to move from naming the subject to naming the work performed.

DimensionReasoningScore

Specificity

The description names the domain and mechanism concretely ("Subdomain takeover via dangling DNS/CNAME") and lists many specific providers ("GitHub Pages, Heroku, Azure, Fastly, Shopify, Netlify..."), but it describes the subject rather than enumerating concrete actions (e.g., enumerate, fingerprint, detect, claim). This matches the anchor 'names domain and 1-2 concrete actions' — the single implicit action is 'takeover' — rather than the level-4 anchor's 'several specific actions', and is well above level 2's 'actions are minimal or generic' given the provider specificity.

3 / 5

Completeness

The 'what' is clear (subdomain takeover via dangling DNS/CNAME across listed providers), but there is no 'Use when...' clause or equivalent explicit trigger guidance in the description field — the trigger phrases sit in a separate metadata.when_to_use field instead. Per the judging guideline, a missing 'Use when...' clause caps completeness at 3, matching the anchor 'has a clear what but when is missing or only weakly implied'.

3 / 5

Trigger Term Quality

Natural phrases users would actually say are present — "subdomain takeover", "dangling DNS", "dangling CNAME" — plus concrete provider names. A few natural terms are missing from the description itself (e.g., "bug bounty", "orphaned subdomain", "CNAME", and tool names like subjack/subzy, which live only in metadata.when_to_use), so it fits anchor 4 ('good keyword coverage; a few natural terms missing') rather than the comprehensive level-5 anchor, while clearly exceeding level 3's 'missing common variations or synonyms'.

4 / 5

Distinctiveness Conflict Risk

The description carves out a clear niche with distinct triggers — "dangling DNS/CNAME" combined with specific provider names — that would rarely conflict with other skills. It matches anchor 5 ('clear niche with distinct triggers; minimal conflict risk') and is more specific than the level-4 example, which is broader ('Works with PDF and Word document files').

5 / 5

Total

15

/

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

metadata_version

'metadata.version' is missing

Warning

Total

15

/

16

Passed

Repository

PurpleAILAB/Decepticon

Path

packages/decepticon/decepticon/skills/standard/recon/web-recon/subdomain-takeover/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.