CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

attacking-oauth-with-device-code-phishing

Run OAuth 2.0 device-code and illicit-consent phishing attacks against Microsoft Entra ID, using TokenTactics-style tooling to steal access and refresh tokens, bypass MFA, and pivot across Microsoft 365 services. Use for authorized red-team engagements simulating device-code or consent-grant phishing against a tenant you have explicit written permission to test.

Quality

87%

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

SKILL.md
Quality
Evals
Security

Quality

Content

71%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 highly actionable, well-sequenced offensive workflow with copy-paste commands and a validation checklist. It loses points for an over-explanatory Overview and for inlining API reference material that already exists in a bundle file without linking to it.

Suggestions

Trim the Overview: drop the general explanation of the device-code grant and the threat-actor name-dropping, keeping only the attack-specific framing red teams need.

Replace inlined endpoint/client-ID/function catalogs in the phases with pointers to references/api-reference.md (e.g., 'See [api-reference.md](references/api-reference.md) for the full client-ID and cmdlet catalog'), and link standards.md for the MITRE/NIST mapping.

Add explicit error-recovery checkpoints for poll responses (expired_token, slow_down, authorization_declined) and for failed refresh-token exchanges, so each phase has a clear feedback loop.

DimensionReasoningScore

Conciseness

Mostly efficient, but the Overview re-explains the OAuth device-code grant (a concept Claude already knows) and devotes a paragraph to threat-actor name-dropping (Storm-2372, Volexity, Proofpoint, Mandiant) that is justification, not actionable guidance.

3 / 5

Actionability

Every phase gives copy-paste-ready, executable commands — curl devicecode/token calls, TokenTactics PowerShell cmdlets, roadtx invocations, real first-party client IDs, and a JWT decode one-liner — covering the common cases concretely.

5 / 5

Workflow Clarity

Six clearly sequenced phases with a token-decode verification checkpoint (Phase 3 step 3) and a terminal Validation Criteria checklist, but lacks explicit error-recovery feedback loops (e.g., handling expired_token / slow_down or a failed refresh).

4 / 5

Progressive Disclosure

The body is well-sectioned into phases, but it inlines substantial API/endpoint/client-ID detail that duplicates references/api-reference.md, and neither api-reference.md nor standards.md is linked or signaled from the body, so navigation to the bundle is unclear.

3 / 5

Total

15

/

20

Passed

Description

100%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.

An excellent description: it states concrete offensive actions, the target platform and tooling, and an explicit authorized-use trigger clause, all in a distinct niche. Verb form matches the rubric's own good examples, so no voice penalty applies.

DimensionReasoningScore

Specificity

Lists multiple concrete actions — 'Run OAuth 2.0 device-code and illicit-consent phishing attacks', 'steal access and refresh tokens', 'bypass MFA', and 'pivot across Microsoft 365 services' — giving comprehensive coverage rather than vague abstraction.

5 / 5

Completeness

Explicitly answers both what ('Run … phishing attacks … steal … tokens, bypass MFA, and pivot across Microsoft 365 services') and when ('Use for authorized red-team engagements simulating device-code or consent-grant phishing against a tenant you have explicit written permission to test').

5 / 5

Trigger Term Quality

Dense with natural domain terms a red-teamer would actually say — 'device-code', 'illicit-consent', 'consent-grant phishing', 'Microsoft Entra ID', 'TokenTactics', 'access and refresh tokens', 'MFA' — including synonyms and tool names.

5 / 5

Distinctiveness Conflict Risk

Occupies a clear narrow niche (device-code / consent-grant phishing against Entra ID via TokenTactics) with distinct triggers and minimal overlap with other skills.

5 / 5

Total

20

/

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

mukul975/Anthropic-Cybersecurity-Skills

Path

skills/attacking-oauth-with-device-code-phishing/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.