cve-hunting-methodology
Skill2 starsChanged 2 months agoMITView on GitHub
Cross-cutting methodology for the CVE-hunting pipeline (ported from find-cve-agent): when a bug is design-vs-real, false-positive avoidance, version checking before reporting, responsible disclosure norms, acceptance-rate expectations per vuln class, vulnerability chaining playbook, known false-positive patterns, maintainer-response patterns, secure-pattern reference (what NOT to flag), and a self-criticism checklist to run before claiming a finding. Use this alongside the cve-hunter/cve-exploiter/cve-validator agents and the /cve-hunt pipeline — load it whenever judging if a finding is real, worth reporting, or already fixed.
In this repository255 other files it gives its agents.
- ad-acl-abuseSkill
- ad-adcs-access-and-relaySkill
- ad-adcs-persistenceSkill
- ad-adcs-template-abuseSkill
- ad-ad-discoverySkill
- ad-ad-persistenceSkill
- ad-auth-coercion-relaySkill
- ad-credential-dumpingSkill
- ad-gpo-abuseSkill
- ad-kerberos-delegationSkill
- ad-kerberos-roastingSkill
- ad-kerberos-ticket-forgingSkill
- ad-pass-the-hashSkill
- ad-sccm-exploitationSkill
- ad-trust-attacksSkill
- hunt-ntlm-infoSkill
- offensive-active-directorySkill
- hunt-llm-aiSkill
- hunt-rag-vectorSkill
- offensive-ai-securitySkill
- cloud-iam-deepSkill
- hunt-cloud-misconfigSkill
- hunt-k8sSkill
- m365-entra-attackSkill
- network-container-escapesSkill
- offensive-cloudSkill
- offensive-container-escapeSkill
- offensive-k8s-attacksSkill
- okta-attackSkill
- vmware-vcenter-attackSkill
- credential-password-sprayingSkill
- hunt-brute-forceSkill
- hunt-forgot-passwordSkill
- post-exploit-credential-recoverySkill
- meme-coin-auditSkill
- offensive-crypto-attacksSkill
- web3-auditSkill
- ctfSkill
- advisory-miningSkill
- auth-bypassSkill
- code-injection-codegenSkill
- command-injectionSkill
- cross-pollinationSkill
- decompression-bombSkill
- entity-expansionSkill
- fp-checkSkill
- method-clobberingSkill
- path-traversalSkill
- prototype-pollutionSkill
- recursion-dosSkill
- redosSkill
- sandbox-escapeSkill
- target-reconSkill
- web2-vuln-classesSkill
- evasion-av-edr-evasionSkill
- offensive-anti-forensicsSkill
- offensive-edr-evasionSkill
- offensive-keylogger-archSkill
- offensive-mitigationsSkill
- offensive-waf-bypassSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 0 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 27
Before the cve-validator agent signs off on a finding → read false-positive-avoidance.md + self-criticism-checklist.md.
- Very specific: line 28
Before picking a target category in /cve-recon → read acceptance-rates.md.
- Very specific: line 29
After confirming one bug → read chaining-playbook.md before stopping.
- Very specific: line 30
Before drafting a disclosure → read disclosure-channels.md + responsible-disclosure.md.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
CVE Hunting Methodology
Reference material for judgment calls during the CVE-hunting pipeline (/cve-recon → /cve-hunt → cve-hunter → cve-exploiter → cve-validator → /cve-registry). This skill has no scripts to run — it's the accumulated judgment layer. Read the relevant reference file at the matching decision point:
Rules (apply before calling anything a finding)
references/design-vs-bug.md— is this actually a bug, or intended behavior/documented limitation?references/false-positive-avoidance.md— the checks to run before writing up a findingreferences/version-checking.md— confirm the vulnerable code path still exists in the current released version, not just in history
Knowledge (context for prioritization and write-ups)
references/acceptance-rates.md— which vuln classes actually get accepted/paid, to prioritize hunting effortreferences/chaining-playbook.md— how to escalate a low-impact finding into something that matters by chainingreferences/false-positive-patterns.md— catalog of common FP shapes per vuln classreferences/rejection-taxonomy.md— why maintainers/triagers reject reports, categorizedreferences/maintainer-responses.md— patterns in how maintainers respond, and how to read themreferences/secure-patterns.md— code patterns that are ALREADY safe — don't flag thesereferences/self-criticism-checklist.md— devil's-advocate pass to run on yourself before submitting a verdictreferences/disclosure-channels.md— picking HackerOne vs GHSA vs direct email vs full disclosurereferences/responsible-disclosure.md— timelines and etiquette for coordinated disclosure
When to use this
- Before the cve-validator agent signs off on a finding → read
false-positive-avoidance.md+self-criticism-checklist.md. - Before picking a target category in
/cve-recon→ readacceptance-rates.md. - After confirming one bug → read
chaining-playbook.mdbefore stopping. - Before drafting a disclosure → read
disclosure-channels.md+responsible-disclosure.md.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

