Skip to content
Agentmart
Submit a tool

zerodayhunt

Use when hunting for zero-days, backdoors, RCE, supply chain attacks, JWT vulnerabilities, cache poisoning, HTTP smuggling, dependency confusion, source map…

Skill by kalpmodi

IndexedMIT★ 23
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/kalpmodi-zerodayhunt

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Use when hunting for zero-days, backdoors, RCE, supply chain attacks, JWT vulnerabilities, cache poisoning, HTTP smuggling, dependency confusion, source map exposure, GSRM/WAF bypass, internal admin panel exposure, business logic flaws, race conditions, subdomain takeover, cloud misconfigs, mobile APK secrets, OAuth attacks, CORS misconfigs, serialization/SSTI/XXE, CI/CD pipeline attacks, or chained attack vectors. Also use when user says "find zero day", "hunt backdoor", "find RCE", "go deep", "maximum potential", "find critical", "chain attack", or "elite hunt".

Use cases: devops, data, security

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/zerodayhunt

What's inside

Skill name
zerodayhunt
Description
Use when hunting for zero-days, backdoors, RCE, supply chain attacks, JWT vulnerabilities, cache poisoning, HTTP smuggling, dependency confusion, source map exposure, GSRM/WAF bypass, internal admin panel exposure, business logic flaws, race conditions, subdomain takeover, cloud misconfigs, mobile APK secrets, OAuth attacks, CORS misconfigs, serialization/SSTI/XXE, CI/CD pipeline attacks, or chained attack vectors. Also use when user says "find zero day", "hunt backdoor", "find RCE", "go deep", "maximum potential", "find critical", "chain attack", or "elite hunt".
License
MIT
Instructions
176 lines in SKILL.md · view on GitHub

Files (21)

  • SKILL.md
  • admin-infra.md · tech/
  • business-logic.md · tech/
  • chain-blueprints.md · tech/
  • cicd.md · tech/
  • client-proto.md · tech/
  • cors-host.md · tech/
  • file-crypto.md · tech/
  • github-js-sourcemap.md · tech/
  • graphql.md · tech/
  • jwt-saml-sso.md · tech/
  • mobile-apk.md · tech/
  • output.md · tech/
  • race-timing.md · tech/
  • ssrf-oob.md · tech/
  • ssti-deser-xxe.md · tech/
  • supply-chain.md · tech/
  • takeover-cloud.md · tech/
  • waf-header-mining.md · tech/
  • websocket-api.md · tech/
  • xs-leaks.md · tech/

SKILL.md (first lines)

# Zero-Day & Critical Vulnerability Hunt

## Philosophy
Automated tools find known CVEs. This skill finds what scanners miss:
logic flaws, architecture leaks, supply chain vectors, auth bypasses,
and chained attacks. Evidence required before claiming ANY finding.
Never hallucinate - if you didn't see it in a response body, it didn't happen.

**The rule that separates elite from script kiddie:** A single finding is rarely Critical.
The payout comes from CHAINING. SSRF + AWS metadata = IAM creds = cloud takeover.
Info disclosure + IDOR = mass PII = Critical. Always ask: "What can I do WITH this?"

**Critical rule:** NEVER stop at "SSRF/network reachability possible" or "endpoint exists".
Must prove full chain: credentials extracted, PII accessed, or data exfiltrated.
A 200 with empty body = WAF catch-all, not a real finding.

## ARGUMENTS
`<target>` - domain (e.g. target.com)
`<focus>` - optional: RCE / SSRF / IDOR / SUPPLY-CHAIN / JWT / BUSINESS-LOGIC / CHAIN / ALL

---

## Phase 0: Smart Intake

```bash
source ~/.claude/skills/_shared/phase0.sh
source ~/.claude/skills/_shared/signals.sh
p0_init_vars "$1"
p0_state_gate "HARVEST" || exit 0
p0_read_relay recon secrets exploit
p0_read_memory
p0_read_hypotheses

# Aliases for downstream phase references
TECH_STACK=$KNOWN_TECH
ALL_ENDPOINTS="$INTERESTING_ENDPOINTS $WAYBACK_ENDPOINTS $API_SPEC_ENDPOINTS"
ATW_FLAGGED=$(echo "$ATW_FLAGGED" 2>/dev/null)

echo "=== ZERODAYHUNT SMART INTAKE: $TARGET ==="
echo "State: $STATE | WAF: $WAF | Tech: $TECH_STACK"
echo "Top hypothesis: $TOP_HYPO_LABEL [$TOP_HYPO_PROB%]"
echo "SSRF vectors from exploit: $(echo "$SSRF_VECTORS" | wc -l)"
echo "Internal IPs available: $(echo "$INTERNAL_IPS" | wc -w)"
echo "JWT tokens: $(echo "$JWT_TOKENS" | wc -l)"
echo "Prior confirmed findings: $(echo "$CONFIRMED_VULNS" | wc -l)"
echo "ATW flagged (skip): $ATW_FLAGGED"
echo ""
echo "Phase selection - run phases where we have targets:"
[ -n "$SSRF_VECTORS" ] && echo "  Phase 8 (SSRF): YES - $(echo "$SSRF_VECTORS" | wc -l) vectors to chain"
[ -n "$JWT_TOKENS" ] && echo "  Phase 7 (JWT): YES - tokens found"
[ -n "$INTERNAL_IPS" ] && echo "  Phase 8 (SSRF pivot): YES - internal IPs to target"
[ -n "$CONFIRMED_VULNS" ] && echo "  Phase 23 (Chain): YES - $(echo "$CONFIRMED_VULNS" | wc -l) confirmed to chain"
```

**Chain-first strategy:** Look at what's already confirmed in prior phases and ask "What can I chain WITH this?" before hunting new vectors. SSRF already confirmed + AWS detected = immediately go to IMDSv1 chain. Auth bypass confirmed = immediately check IDOR chain.

## Phase 1 - Full Context Build

- `SSRF_VECTORS` -> target these for cloud metadata chains
- `INTERNAL_IPS` -> SSRF target list for lateral movement
- `JWT_TOKENS` -> JWT confusion candidates (check alg header)
- `TECH_STACK` -> picks specific phases to prioritize (see table below)
- `VERIFIED_CREDS` -> credential spray on all surfaces before hunting new vulns
- `WAF` -> Phase 2 first if WAF present
- `ALL_ENDPOINTS` -> full attack surface for every subsequent phase

**Phase priority by tech stack:**

| Tech detected | Run these phases FIRST |
|---|---|
| AWS + SSRF_VECTOR | Phase 8 -> Phase 12 (cloud asset chain) |
| Node.js + JWT | Phase 7 (JWT confusion), Phase 25 (prototype pollution) |
| Spring Boot | Phase 15 (SSTI/deserialization), Phase 16 (CI/CD) |
| OAuth/OIDC | Phase 7 (JWT), Phase 31 (SAML/SSO) |
| GraphQL | Phase 17 (introspection, batching) |
| React/webpack | Phase 18 (source maps), Phase 5 (JS bundles) |
| any | Phase 23 (chain blueprints) - always run this last |

---

## Execution Manifest

Write to `session.json.scalpel.active_manifest` before starting:

```json
{
  "skill": "zerodayhunt",
  "target": "<target>",
  "items": [
    {"id": "zdh01", "priority": "MUST",     "tool": "waf-header-mining",  "target": "<target>", "status": "pending"},
    {"id": "zdh02", "priority": "MUST",     "tool": "github-js-sourcemap","target": "<target>", "status": "pending"},
    {"id": "zdh03", "priority": "MUST",     "tool": "ssrf-oob",           "target": "<target>", "status": "pending"},
    {"id": "zdh04", "priority": "MUST",     "tool": "jwt-saml-sso",       "target": "<target>", "status": "pending"},
    {"id": "zdh05", "priority": "MUST",     "tool": "chain-blueprints",   "target": "<target>", "status": "pending"},
    {"id": "zdh06", "priority": "SHOULD",   "tool": "business-logic",     "target": "<target>", "status": "pending"},
    {"id": "zdh07", "priority": "SHOULD",   "tool": "supply-chain",       "target": "<target>", "status": "pending"},
    {"id": "zdh08", "priority": "SHOULD",   "tool": "admin-infra",        "target": "<target>", "status": "pending"},
    {"id": "zdh09", "priority": "SHOULD",   "tool": "ssti-deser-xxe",     "target": "<target>", "status": "pending"},
    {"id": "zdh10", "priority": "SHOULD",   "tool": "cors-host",          "target": "<target>", "status": "pending"},
    {"id": "zdh11", "priority": "SHOULD",   "tool": "race-timing",        "target": "<target>", "status": "pending"},
    {"id": "zdh12", "priority": "SHOULD",   "tool": "cicd",               "target": "<target>", "status": "pending"},
    {"id": "zdh13", "priority": "SHOULD",   "tool": "graphql",            "target": "<target>", "status": "pending"},
    {"id": "zdh14", "priority": "IF_TIME",  "tool": "takeover-cloud",     "target": "<target>", "status": "pending"},
    {"id": "zdh15", "priority": "IF_TIME",  "tool": "mobile-apk",         "target": "<target>", "status": "pending"},
    {"id": "zdh16", "priority": "IF_TIME",  "tool": "client-proto",       "target": "<target>", "status": "pending"},
    {"id": "zdh17", "priority": "IF_TIME",  "tool": "file-crypto",        "target": "<target>", "status": "pending"},
    {"id": "zdh18", "priority": "IF_TIME",  "tool": "websocket-api",      "target": "<target>", "status": "pending"},
    {"id": "zdh19", "priority": "IF_TIME",  "tool": "xs-leaks",           "target": "<target>", "status": "pending"},
    {"id": "zdh20", "priority": "MUST",     "tool": "output",             "target": "<target>", "status": "pending"}
  ]
}
```

---

## Technique Loader

Load the relevant technique file from `~/.claude/skills/zerodayhunt/tech/` based on active manifest item:

From kalpmodi/akira, MIT.

More from kalpmodi/akira

Signals

Stars
23
Forks
4
Open issues
0
Last commit (this folder)
Apr 25, 2026 (6 months ago)
Last release
Apr 25, 2026 (6 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from kalpmodi

Cite this page

zerodayhunt skill. Agent Mart. https://agentmart.dev/skills/kalpmodi-zerodayhunt (last updated 2026-10-24)