Skip to content
ai-supply.store
DécouvrirCatégoriesClassementsCommunautéAgent APIFAQ
Se connecterInscription gratuite
catalog / Language & NLP / Stanford Alpaca
▣DatasetLanguage & NLPFree

Stanford Alpaca

Code and the 52K instruction-following dataset behind Stanford's Alpaca models.

@ai-supply
Installations30k
↗ Dépôt source
← More Language & NLPLanguage & NLP leaderboard →How we grade security →Source ↗

Stanford Alpaca

Code, documentation, and the instruction-following dataset behind Stanford's Alpaca models, which fine-tuned LLaMA on 52K instruction–response pairs generated via self-instruct.

The repository includes the dataset, the data-generation pipeline, and training details, and was highly influential in the open instruction-tuning wave.

Apache-2.0 licensed; a foundational reference dataset for instruction tuning and alignment research.

Rating rank
#1
of 30 in Language & NLP
Install rank
#20
of 30 in Language & NLP
Security score
75/100 · B
review
Security rank
#15
of 30 in Language & NLP
Installs
30k
cat avg 145k
This listing vs category average
Installs
this
cat avg
Security (of 100)
this
cat avg
Adoption trend
See the Language & NLP leaderboard →
! Security: Review · 7575/100 · grade Bscanned 15h ago
✓ no compromise signals6 risk-surface · 8/20 OWASP controls flagged

Compromise signals — malicious or tampered code (leaked secrets, backdoors, a dropped executable) — reduce the score, and known dependency CVEs carry a bounded penalty (they warrant review but never QUARANTINE — update the dependency to clear). Other dangerous-by-capability traits are risk surface, expected for some capabilities. Every finding is mapped to its OWASP control below.

Data card · high confidence (static)
jsontxtjsonl
PII surface: Email addresses present, Phone number present
22 files

Findings mapped to the OWASP Top 10 for LLM Applications (2025) and the OWASP Machine Learning Security Top 10. Expand any flagged control for the exact findings — compromise reduces the score; expected/risk-surface do not.

OWASP Top 10 for LLM Applications
⚠LLM03Supply Chaincritical
Vulnerable/compromised dependencies, models or archives in the artifact.
•Vulnerable dependencies — 35 known vulnerabilities in: transformers@4.28.1 (CWE-1395)known CVE · -25 pts
⚠LLM02Sensitive Information Disclosuremedium
Secrets, credentials or PII shipped inside the artifact.
•Email addresses present — contains email-like strings · tatsu-lab-stanford_alpaca-761dc5b/seed_tasks.jsonlrisk surface
•Phone number present — contains phone number-like pattern (E.164 or formatted) · tatsu-lab-stanford_alpaca-761dc5b/seed_tasks.jsonl (CWE-359)risk surface
⚠LLM08Vector and Embedding Weaknessesmedium
PII or plaintext source leakage in embedding/vector exports.
Embedding inversion/poisoning is largely runtime; static check covers PII in vector exports.
•Email addresses present — contains email-like strings · tatsu-lab-stanford_alpaca-761dc5b/seed_tasks.jsonlrisk surface
•Phone number present — contains phone number-like pattern (E.164 or formatted) · tatsu-lab-stanford_alpaca-761dc5b/seed_tasks.jsonl (CWE-359)risk surface
⚠LLM10Unbounded Consumptionmedium
Unbounded loops/recursion causing DoS or runaway cost.
Enforced at runtime by the gateway (rate limits + spend caps + size caps); static check flags unbounded loops.
•Potentially unbounded loop — an infinite loop (while True / while(1) / for(;;)) may cause runaway consumption · tatsu-lab-stanford_alpaca-761dc5b/utils.py (CWE-835)risk surface
⚠LLM06Excessive Agencylow
Over-broad tool/permission surface or unrestricted egress.
•Broad capability surface — 3 high-impact capability categories referenced — verify least-privilege · tatsu-lab-stanford_alpaca-761dc5b/seed_tasks.jsonl (CWE-272)risk surface
§LLM09MisinformationGovernance
Artifacts designed to produce false/deceptive output.
Detectable only by runtime behavioral evaluation; addressed via responsible-use attestation.
✓LLM01Prompt InjectionPassed
✓LLM04Data and Model PoisoningPassed
Backdoors/poisoning in training data or serialized models.
Behavioral poisoning needs model execution; static check covers unsafe serialization + dataset skew only.
✓LLM05Improper Output HandlingPassed
✓LLM07System Prompt LeakagePassed
OWASP Machine Learning Security Top 10
⚠ML06AI Supply Chaincritical
Compromised PyPI/npm packages, typosquats, unsafe serialized models.
•Vulnerable dependencies — 35 known vulnerabilities in: transformers@4.28.1 (CWE-1395)known CVE · -25 pts
⚠ML02Data Poisoningmedium
Poisoned training datasets with triggers or anomalous distributions.
Static check covers trigger phrasing, PII and label skew; full poisoning detection is runtime.
•Email addresses present — contains email-like strings · tatsu-lab-stanford_alpaca-761dc5b/seed_tasks.jsonlrisk surface
•Phone number present — contains phone number-like pattern (E.164 or formatted) · tatsu-lab-stanford_alpaca-761dc5b/seed_tasks.jsonl (CWE-359)risk surface
⚠ML05Model Theftlow
Unlicensed re-distribution / license-incompatible derivatives.
Static check verifies license declaration; extraction throttling is runtime.
•No license signal — no SPDX id or license keyword found · tatsu-lab-stanford_alpaca-761dc5b/.gitignorerisk surface
§ML01Input Manipulation (Adversarial)Governance
Models vulnerable to adversarial perturbations.
Requires runtime robustness evaluation; addressed via publisher robustness attestation.
§ML03Model InversionGovernance
Training data reconstructable from a model's outputs.
Runtime/evaluation property; addressed via model-card data-provenance + DP attestation.
§ML04Membership InferenceGovernance
Determining whether a record was in the training set.
Runtime/evaluation property; addressed via overfitting disclosure + DP attestation.
§ML08Model SkewingGovernance
Models trained on skewed data producing biased output.
Requires fairness evaluation; addressed via model-card bias/limitations disclosure.
◷ML09Output IntegrityRuntime-enforced
Middleware tampering with model outputs in transit.
Gateway enforces TLS + response integrity; static check flags output-rewriting code.
✓ML07Transfer Learning AttackPassed
Backdoored base models / LoRA adapters propagating to derivatives.
Backdoor detection needs behavioral probing; static check covers unsafe serialization + provenance.
✓ML10Model Poisoning (Weights)Passed
Tampered model weight files; integrity must be verifiable.
Static check enforces safe formats + records a content hash for downstream verification.
Other findings (4) · hygiene / uncategorized
•Unrecognized file type — '.gitignore' is not on the allowlist · tatsu-lab-stanford_alpaca-761dc5b/.gitignorerisk surface
•Unrecognized file type — '.?' is not on the allowlist · tatsu-lab-stanford_alpaca-761dc5b/DATA_LICENSErisk surface
•Unrecognized file type — '.jsonl' is not on the allowlist · tatsu-lab-stanford_alpaca-761dc5b/seed_tasks.jsonlrisk surface
•Possible obfuscation — very long lines · tatsu-lab-stanford_alpaca-761dc5b/seed_tasks.jsonlrisk surface
✔ verified source · pinned tatsu-lab-stanford_alpaca-761dc5b
Check against a policy

The same gate an agent runs before installing (POST /api/v1/trust/stanford-alpaca/check). Click a policy:

Consume Stanford Alpaca programmatically. Authenticate with an API key or session — see Authorize an agent.

# Agents: CHECK BEFORE YOU INSTALL (no auth) — score, grade, level, capability manifest
curl https://ai-supply.store/api/v1/trust/stanford-alpaca

# Gate against your org policy (returns { pass, violations })
curl -X POST https://ai-supply.store/api/v1/trust/stanford-alpaca/check \
  -H "Content-Type: application/json" \
  -d '{"minGrade":"B","denyPermissions":["shell"],"denyUnknownEgress":true}'

# CLI
npx ai-supply add stanford-alpaca

# REST (install → download)
curl -X POST https://ai-supply.store/api/v1/listings/stanford-alpaca/install \
  -H "Authorization: Bearer $AIM_KEY"

# MCP tool
install_listing({ "slug": "stanford-alpaca" })
OpenAPI spec →
vlatest
! Security: Review · 7517h ago

Curated mirror — latest upstream source. See the repository for tagged releases.

Sign in and install this listing to leave a review.

More from @ai-supply

View profile →
◉Agent
MetaGPT
Multi-agent framework that assigns GPT roles (PM, engineer, QA) to solve complex software tasks end-to-end.
↓ 1.0M
⇄Connector
vLLM
High-throughput, memory-efficient LLM inference engine with PagedAttention and continuous batching.
↓ 892k
⇄Connector
Meilisearch
Lightning-fast open-source search engine with typo-tolerance, semantic hybrid search, and sub-50ms response times.
↓ 811k
△Eval
Weights & Biases (wandb)
ML experiment tracking and visualization — log metrics, hyperparameters, models, and media in real time.
↓ 784k
ai-supply.store

Des capacités d'IA gratuites et vérifiées pour la sécurité — skills, MCP, plugins, agents, datasets et bien plus, chacune notée et suivie pour rester à jour, et pensée autant pour les humains que pour les agents.

api · v3.1status · all green
Contact
support@ai-supply.storesecurity@ai-supply.store
Catalogue
  • Découvrir
  • Catégories
  • Classements
  • Benchmarks
  • Sécurité
Communauté
  • Communauté
  • FAQ
Pour les agents
  • Démarrage rapide (60s)
  • Autoriser un agent
  • Agent API
  • Spécification OpenAPI
Pour les développeurs
  • Publier
  • Tableau de bord
Compte
  • Créer un compte
  • Se connecter
  • Paramètres
Mentions légales
  • Conditions
  • Accord éditeur
  • Utilisation acceptable
  • Confidentialité