hunt-race-condition

Hunting skill for race condition vulnerabilities. Built from 12 public bug bounty reports including modern HTTP/2 single-packet attack cases (James Kettle DEF CON 2023 "Smashing the State Machine"; RyotaK / Flatt Security 10,000-request first-sequence-sync expansion 2024). Covers coupon double-redemption, gift-card double-spend, MFA-OTP-validate race, account-create race, faucet/crypto token double-mint, email-activation race, vote/upvote inflation, password-reset token race, rate-limit bypass via concurrent requests. Use when hunting race conditions, TOCTOU bugs, MFA-bypass-via-timing.

Quality

83%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Critical

Do not install without reviewing

Quality

Content

77%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.

Highly actionable content with executable tooling and a well-validated hunting workflow, but it is a verbose monolith that repeats the single-packet technique across sections and inlines deep-reference and citation material that would benefit from separate bundle files.

Suggestions

Move the HTTP/2 Single-Packet Deep Reference and the Disclosed Report Citations into separate files under references/ and keep SKILL.md as a concise overview that links one level deep, which would improve both conciseness and progressive_disclosure.

De-duplicate the single-packet explanation: keep one canonical description (the Deep Reference) and have the earlier Firing-a-race section briefly point to it instead of re-explaining the technique.

Tighten the Real Impact Examples — three full prose scenarios restate the cited cases; condensing to one-liners that reference the citation numbers would cut tokens without losing signal.

Dimension	Reasoning	Score
Conciseness	Mostly efficient specialist content, but the ~485-line body explains the HTTP/2 single-packet attack at three different depths (Firing a race, the citation list, and the Deep Reference) and the three verbose Real Impact scenarios largely restate the cited cases, so it could be tightened.	3 / 5
Actionability	Provides copy-paste-ready, executable artifacts throughout — a complete line-by-line Turbo Intruder BURP2 template, curl/asyncio/grep snippets, tcpdump+Wireshark validation commands, and a concrete 7-step race-window estimation methodology covering the common cases.	5 / 5
Workflow Clarity	A 10-step numbered hunting methodology with explicit validation checkpoints ("Analyze responses", "Verify the effect", reproducibility over 5 attempts) plus a Gate 0 validation section that gates reporting on demonstrated harm and reproducibility — clear sequence with feedback loops and a checklist.	5 / 5
Progressive Disclosure	Good in-file section headers give structure, but with no bundle files present the whole skill is a monolith: the Deep Reference and the 9 backfilled citations clearly belong in separate reference files, so content that should be split is inlined.	3 / 5
	Total	16 / 20 Passed

Description

88%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 strong, specific description that comprehensively enumerates race-condition subclasses and pairs a clear what with an explicit Use-when clause. Minor overlap risk with adjacent hunting skills and slightly jargon-heavy trigger phrasing keep it just short of perfect.

Dimension	Reasoning	Score
Specificity	Lists multiple specific concrete race subclasses — "coupon double-redemption, gift-card double-spend, MFA-OTP-validate race, account-create race, faucet/crypto token double-mint, email-activation race, vote/upvote inflation, password-reset token race, rate-limit bypass" — giving comprehensive coverage of the domain.	5 / 5
Completeness	Explicitly answers both what (a hunting skill covering the enumerated race subclasses) and when ("Use when hunting race conditions, TOCTOU bugs, MFA-bypass-via-timing") with concrete trigger phrases.	5 / 5
Trigger Term Quality	Includes natural phrases a user would say ("race conditions", "TOCTOU bugs", "hunting race conditions"), but "MFA-bypass-via-timing" and the cited researcher names are more jargon than natural trigger language, leaving a few common synonyms (e.g. "concurrency bugs") uncovered.	4 / 5
Distinctiveness Conflict Risk	The race-condition niche is clear and mostly distinct, but the "MFA-OTP-validate race" / "MFA-bypass-via-timing" triggers create minor overlap risk with sibling skills like hunt-mfa-bypass and hunt-business-logic.	4 / 5
	Total	18 / 20 Passed

Validation

93%

Warnings & errors only

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

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: elementalsouls/Claude-BugHunter
Path: skills/hunt-race-condition/SKILL.md
Commit: 58017df

Reviewed: 3 days ago

Table of Contents

Discovery Implementation Validation

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.