SkillCheatSheet
detecting-data-and-model-poisoning
An AI agent skill from mukul975/Anthropic-Cybersecurity-Skills, scored 63 out of 100 by code.
Code score by group
It loses the most points in Procedure (15), the group that checks whether the steps are numbered, read the context first and say what to do when something fails.
What the author says it does
Identify poisoned training data and backdoored ML models across the pipeline using IBM's Adversarial Robustness Toolbox (activation clustering, spectral signatures, trigger reconstruction), Cleanlab for label-quality issues, and supply-chain checks like weight-hash verification and safetensors enforcement. Use before training or deploying on third-party/user-contributed data or downloaded checkpoints, during ML supply-chain reviews, or when investigating model misbehavior tied to specific inputs (suspected backdoor trigger).
This is the description in the skill's own SKILL.md, written by its author, not by SkillCheatSheet.
- Repository
- mukul975/Anthropic-Cybersecurity-Skills
- Folder
- skills/detecting-data-and-model-poisoning
- License
- Apache-2.0
- GitHub stars
- 35,958
- SKILL.md lines
- 216
- Other files
- 4
Add detecting-data-and-model-poisoning to your agent
npx skills add mukul975/Anthropic-Cybersecurity-Skills --skill detecting-data-and-model-poisoningThe command uses the open skills CLI. A skill is instructions and scripts your agent will follow, so read it first: open the folder on GitHub at the commit that was scored.
More skills from mukul975/Anthropic-Cybersecurity-Skills
-
exploiting-adcs-with-certipy
Use Certipy to enumerate AD CS certificate authorities and templates over LDAP/RPC, then exploit ESC1-ESC16 misconfigurations - SAN abuse, NTLM relay to web enrollment (ESC8), Sha…
Verified84B -
auditing-tls-certificate-transparency-logs
Monitors Certificate Transparency (CT) logs to detect unauthorized certificate issuance, discover subdomains via CT data, and alert on suspicious certificate activity for owned do…
Verified82B -
conducting-api-security-testing
Conducts security testing of REST, GraphQL, and gRPC APIs to identify vulnerabilities in authentication, authorization, rate limiting, input validation, and business logic. The te…
Verified82B -
conducting-gdpr-compliance-assessment
Conduct comprehensive GDPR compliance assessments by evaluating data processing activities against EU Regulation 2016/679, including Article 30 records of processing, lawful basis…
Verified82B -
detecting-container-runtime-threats-with-falco
Deploys and operates Falco with the modern eBPF driver in Kubernetes and Docker, covering driver selection, Helm installation, output channels, and the built-in ruleset that detec…
Verified82B -
detecting-typosquatting-packages
Flag misspelled, brandjacked, and typosquatted package names across npm, PyPI, and crates.io before installation, using edit-distance, keyboard-proximity, and known-target corpus…
Verified82B -
performing-access-review-and-certification
Designs and runs access review and certification campaigns-scoping, reviewer selection, risk-based prioritization, micro-certification, and remediation tracking-to verify user acc…
Verified82B -
assessing-vector-and-embedding-weaknesses
Test RAG vector stores (Pinecone, Qdrant, Weaviate, Chroma, pgvector, FAISS) for embedding inversion, cross-tenant data leakage, and data poisoning per OWASP LLM08:2025. Use when…
Verified81B
All skills in mukul975/Anthropic-Cybersecurity-Skills
Scores come from a snapshot of 300 public GitHub repositories taken on 12 November 2026. Every score is the result of fixed checks run by code; no AI is involved. How scoring works.