/Catalogue/Prompt/elementalsouls/elementalsouls-claude-bughunter-hunt-exceptional-conditions

Origin: github

hunt-exceptional-conditions

Hunt mishandling of exceptional conditions — feed an endpoint malformed/unexpected input (wrong type, broken JSON, oversized field, null byte) and make it fail OPEN or leak internals: a verbose stack-trace / framework error page that discloses ORM internals, server file paths, library versions, or a language traceback. Use on any input-accepting endpoint (JSON APIs, forms, query params). Medium-High when the leak exposes internal structure that arms a deeper attack.

by elementalsouls · updated 5h ago · imported from GitHub

Installs0+0/7d
Security score100/100
Retention 14d0%
GitHub stars4.7K

Skill logic

Execution graph
User message
Prompt rewrites behaviour
Response

SKILL.md

View on GitHub ↗

HUNT-EXCEPTIONAL-CONDITIONS — Verbose Errors / Fail-Open (A10:2025)

What actually pays

Well-built apps catch errors and return a clean, generic message. A broken app, when handed input it didn't expect, throws an unhandled exception and renders a developer error page straight to the client — leaking the stack trace, the ORM/query internals, server-side file paths, and framework/library versions. That disclosure is the finding (and it arms SQLi/RCE/path attacks next).

Recon

Any endpoint that parses input is a candidate; the richest are:

JSON APIs that expect typed fields:  POST /api/* with {numbers, ids, enums}
Endpoints with numeric/id path or query params:  /item/{id}, ?page=, ?quantity=
Search / filter / sort params
File or content-type sensitive uploads

Attack — send what the code didn't anticipate

Take a known-good request and break ONE assumption at a time:

  • Wrong type: a field the app expects to be a number/string is sent as an array or object — {"rating":"x","comment":[1,2,3]}, {"quantity":{}}.
  • Malformed body: truncated/!invalid JSON, an unterminated string, a stray brace, a wrong/missing Content-Type.
  • Boundary/oversized: a very long string, a huge/negative/overflow number.
  • Null byte / control chars embedded in a value.
POST /api/Feedbacks   {"rating":"notanumber","comment":[1,2,3]}
GET  /item/' OR /item/%00   (also exercises the error path)

Watch the RESPONSE BODY, not just the status: a 500 (or even a 200/400) whose body contains a stack trace or framework error page is the signal.

What counts as a leak (the success signal)

A finding is confirmed when the response body contains a cross-framework error-disclosure signature:

  • Node/Express + Sequelize: SequelizeDatabaseError, node_modules/sequelize, a JS stack with internal paths.
  • PHP: <b>Warning</b> ... /var/www/.../file.php on line N.
  • Python: Traceback (most recent call last), werkzeug.exceptions.
  • Java: at com.app.Foo(Foo.java:42) stack frames.
  • .NET: Server Error in '/' Application, a [System.XxxException: ...] YSOD.

A clean JSON error ({"error":"Invalid input"}) with no internals is NOT a finding — that's correct handling. Disclosure of internal structure is.

Validation discipline

  • Capture the exact leaked artifact (path, ORM class, version, stack frame) — that's the evidence. "It returned 500" alone is not disclosure.
  • Note what the leak enables next (e.g. a disclosed SQL error → hunt-sqli; a disclosed absolute path → hunt-lfi).

Discussion

No comments yet — start the thread.

Sign in to join the discussion.

/More from elementalsouls/Claude-BugHunter

elementalsouls· 5h agoCommunity
hunt-business-logic

Prompts · Python · v0.1.0

Hunting skill for business logic vulnerabilities. Built from 12 public bug bounty reports. Covers coupon-race-stacking (Instacart, Stripe, Reverb), negative-quantity-in-cart price tampering (Upserve, Eternal/Zomato), decimal/fraction price-field overflow (Shipt), client-side checkout amount trust on PayPal redirect (WordPress.org), price-per-unit mass-assignment (Krisp), and archived-price swap / cart-TOCTOU (Stripe). Use when hunting business logic — heavy emphasis on financial-impact-demonstrated cases.

#ai-security#anthropic#application-security

0 4.7K
elementalsouls· 5h agoCommunity
hunt-cache-poison

Prompts · Python · v0.1.0

Hunting skill for cache poison vulnerabilities. Built from 10 public bug bounty reports including X-Forwarded-Host poisoning, X-HTTP-Method-Override / GCS cache, reflected→stored XSS via cache, classic Omer-Gil Web Cache Deception, Cloudflare Cache Deception Armor bypass, session-token cache deception, Akamai hop-by-hop smuggling → server-side edge poisoning, and Kettle's 2024 path-normalization WCD against Cloudflare/Fastly/GCP. Host/X-Forwarded-Host injection that reaches app logic (reset-link poisoning, routing SSRF, OAuth issuer) is owned by hunt-host-header; this skill owns the case where the poisoned response is CACHED and served to other users. Use when hunting cache poisoning, Web Cache Deception, CDN-fronted apps.

#ai-security#anthropic#application-security

0 4.7K
elementalsouls· 5h agoCommunity
hunt-captcha-bypass

Prompts · Python · v0.1.0

Hunt CAPTCHA Bypass — 6 distinct patterns: (1) CAPTCHA field simply omitted from the request (server-side validation absent), (2) CAPTCHA token replayed from a solved challenge (no single-use enforcement), (3) CAPTCHA response accepted on a different endpoint than it was solved on (no binding to action/session), (4) static or predictable CAPTCHA values accepted (e.g. '0', 'null', empty string), (5) audio/accessibility CAPTCHA trivially solvable programmatically, (6) CAPTCHA only enforced after N failures (first N requests bypass it). Detection: intercept a successful form submission, remove the CAPTCHA field entirely, replay — if it still succeeds, server-side validation is absent. Medium severity standalone; High when it removes the only rate-limit gate protecting a login, registration, or payment endpoint.

#ai-security#anthropic#application-security

0 4.7K
elementalsouls· 5h agoCommunity
hunt-clickjacking

Prompts · Python · v0.1.0

Hunt Clickjacking — missing X-Frame-Options / CSP frame-ancestors lets an attacker embed the target page in an invisible iframe and trick victims into clicking buttons they cannot see (UI redressing). Targets: login flows, money transfers, account settings, OAuth confirmation pages. Confirm by fetching the page, then PROVE it frames in a real browser and a sensitive state-changing action survives the cross-site context (SameSite cookies / framebusting JS can defeat it) — header-absence alone is not a finding.

#ai-security#anthropic#application-security

0 4.7K