mirror of
https://github.com/mukul975/Anthropic-Cybersecurity-Skills.git
synced 2026-06-26 11:44:37 +03:00
mukul975
8cae0648ec
Demand-driven expansion targeting the fastest-growing 2025-2026 threat and
skills categories (ISC2/WEF/CrowdStrike/Mandiant signals):
- AI Security (NEW domain, 12 skills): LLM red-teaming with garak/PyRIT,
prompt injection (direct/indirect/RAG), MCP tool-poisoning, agentic tool
invocation, guardrails, model/data poisoning, system-prompt leakage,
embedding/vector weaknesses, model extraction, continuous red-teaming
- Supply Chain Security (NEW domain, 5 skills): SBOMs, dependency confusion,
malicious-npm triage, typosquatting, SLSA/Sigstore provenance
- Hardware & Firmware Security (NEW domain, 4 skills): CHIPSEC/UEFI audit,
Secure Boot bypass, TPM measured-boot attestation, ESP bootkit hunting
- Identity (10): Entra ID/ROADtools, GraphRunner, AADInternals, ADCS/Certipy,
shadow credentials, coercion, BloodHound CE, device-code phishing, SSO abuse
- Cloud-native (8): Stratus, Pacu, CloudFox, container escape, K8s RBAC,
Falco, Trivy, kube-bench
- Offensive C2 (6): Sliver, Havoc, NetExec, DPAPI, NTLM relay ESC8, redirectors
- DFIR (6): Hayabusa, Chainsaw, KAPE, Velociraptor, EZ Tools, Plaso
- Backfill (4): OpenCTI, MISP, honeytokens, post-quantum crypto migration
Each skill follows the repo taxonomy (SKILL.md + references/{standards,api-reference}.md
+ scripts/agent.py + LICENSE), with researched real tool commands (no placeholders),
complete frontmatter, and ATT&CK/ATLAS + NIST CSF mappings. Updates README domain
table, skill count, and index.json.
2.3 KiB
2.3 KiB
API and Command Reference
sentence-transformers (embedding generation)
| Call | Purpose |
|---|---|
SentenceTransformer("all-MiniLM-L6-v2") |
Load an embedding model (384-dim) |
model.encode([texts]) |
Return numpy array of embeddings |
model.encode(text, normalize_embeddings=True) |
L2-normalized vectors (for cosine) |
scikit-learn similarity
| Call | Purpose |
|---|---|
cosine_similarity(a, b) |
Pairwise cosine similarity matrix |
Qdrant client (qdrant-client)
| Call | Purpose |
|---|---|
QdrantClient(url="http://localhost:6333") |
Connect |
client.get_collection(name) |
Inspect vector size + distance metric |
client.count(name) |
Corpus size |
client.search(collection_name, query_vector, limit, query_filter) |
k-NN search with optional filter |
client.upsert(name, points=[PointStruct(id, vector, payload)]) |
Insert/update points |
Filter(must=[FieldCondition(key, match=MatchValue(value))]) |
Metadata filter (tenant isolation) |
Chroma (chromadb)
| Call | Purpose |
|---|---|
chromadb.Client() / PersistentClient(path) |
Connect |
collection.query(query_embeddings=[...], n_results=k, where={...}) |
k-NN with metadata filter |
collection.add(ids, embeddings, metadatas, documents) |
Insert |
Pinecone (pinecone-client)
| Call | Purpose |
|---|---|
Pinecone(api_key=...) |
Connect |
index.query(vector=..., top_k=k, namespace="tenant", filter={...}) |
k-NN; namespace = tenant boundary |
index.upsert(vectors=[(id, vec, meta)], namespace=...) |
Insert |
Assessment metrics
| Metric | Meaning |
|---|---|
| Inversion cosine | Similarity between reconstructed candidate and target vector; high = recoverable. |
| Membership delta | top-1 score(in-corpus query) − top-1 score(control query); large positive = membership leak. |
| Poison dominance | Fraction of unrelated queries returning the poison chunk in top_k. |
| Cross-tenant count | Number of foreign-tenant rows returned to a tenant query (should be 0). |
vec2text (research baseline)
| Call | Purpose |
|---|---|
vec2text.load_pretrained_corrector("gtr-base") |
Load inversion corrector for compatible embedder |
vec2text.invert_embeddings(embeddings, corrector) |
Reconstruct text from embeddings |