Skip to content
Cloud Security DeskSearch
Menu

Evidence search

Search

Search titles, summaries, topics, providers, authors, and the full open-access corpus.

Results for “Provenance”

21 publications
Technical guideSource-based analysis

Design application authorization before writing Cedar policies

Define business actions, trustworthy entities and tenant boundaries before writing Cedar policies, then make the application responsible for enforcing the resulting decision.

Identity & access · AWS / Cedar · By Cloud Security Desk
Technical guideSource-based analysis

Make PostgreSQL point in time recovery reproducible

Build a version-aware recovery chain from protected base backups and WAL through timeline selection, isolated replay and application acceptance.

Resilience · PostgreSQL · By Cloud Security Desk
Technical guideSource-based analysis

Choose who can share an inference prefix cache

Choose the principals allowed to share prefix state, then carry that decision through request routing, offload, transfer and restore.

AI systems · vLLM / NVIDIA · By Cloud Security Desk
Technical guideSource-based analysis

Read VPC Flow Logs without overclaiming network evidence

Interpret aggregation, address translation, skipped records and delivery clocks before turning a VPC flow record into an incident conclusion.

Detection & response · AWS · By Cloud Security Desk
Research noteSource-based analysis

Embeddings still need a sensitive data boundary

Treat vector access as a derived-data release decision, with separate permissions for retrieval, debugging and bulk export.

AI systems · By Cloud Security Desk
Technical guideSource-based analysis

Keep build credentials out of the image and its evidence

A temporary BuildKit secret mount controls credential delivery, not everything a build command can do with the credential or leave in its outputs.

Workload security · Docker / GitHub · By Cloud Security Desk
Technical guideSource-based analysis

Give persistent agent memory an admission boundary

Separate conversational candidates from admitted memory, preserve their permitted audience, and make derived summaries repairable.

AI systems · LangChain · By Cloud Security Desk
Technical guideSource-based analysis

Isolate document parsing before RAG ingestion

Give document parsing a bounded worker, then admit its extracted content separately before embedding or indexing.

AI systems · Apache Tika / Kubernetes / gVisor · By Cloud Security Desk
Technical guideSource-based analysis

Keep GitHub audit streaming continuous across maintenance

Plan audit-stream maintenance around native history, pause buffers, receiver acceptance and duplicate-aware evidence receipts.

Detection & response · GitHub / AWS · By Cloud Security Desk
Technical guideSource-based analysis

Workload federation across clouds needs separate trust decisions

Cross-cloud federation adds trust decisions, not a portable permission system. Map each credential and resource grant before moving a runtime workload.

Identity & access · AWS / Google Cloud / Microsoft Entra · By Cloud Security Desk
Technical guideSource-based analysis

A defensible cloud patch queue starts with exploitation evidence

Join exploitation evidence to affected assets, exposure, ownership and verified remediation without turning CVSS, EPSS or a catalog entry into a complete risk score.

Detection & response · CISA / NIST / FIRST · By Cloud Security Desk
Technical guideSource-based analysis

Security evidence for AI release decisions

A release approval should identify the changed application, the claims its tests support and the evidence that expires when a model, prompt, data path or runtime changes.

AI systems · By Cloud Security Desk
Technical guideSource-based analysis

What cloud snapshots cannot preserve

Separate live execution, persisted disk and provider records before an authorized containment decision destroys evidence that a snapshot cannot recover.

Detection & response · AWS / NIST · By Cloud Security Desk
Technical guideSource-based analysis

Testing Sigma detections before a backend change

Keep rule structure, field transformations, query conversion and event matching as separate checks when changing a Sigma detection backend.

Detection & response · SigmaHQ · By Cloud Security Desk
Technical guideSource-based analysis

Where fine tuning data needs a trust boundary

A training dataset can preserve its checksum and still teach the wrong behavior. Admission controls need to separate origin, transformation, approved use and the model change they produce.

AI systems · By Cloud Security Desk
Technical guideSource-based analysis

Turn SBOM and VEX records into patch decisions

A VEX statement is an assertion about a specific product and vulnerability. Match its scope and conditions before using it to suppress a finding.

Workload security · CISA / OpenVEX / CycloneDX / OASIS · By Cloud Security Desk
Technical guideSource-based analysis

The telemetry collector is part of the evidence boundary

Review sender identity, tenant routing, processing and export as separate trust boundaries before treating collected telemetry as dependable evidence.

Detection & response · OpenTelemetry · By Cloud Security Desk
Technical guideSource-based analysis

Put enforceable boundaries around agent tool calls

Treat model proposals as requests for authority, then check the operation, resource, recipient, and approval at the point where a tool can create a side effect.

AI systems · AWS · By Cloud Security Desk
Technical guideSource-based analysis

Verify container provenance before admitting a pinned image

Use the digest to identify the artifact, then check who signed it, which builder produced it, and which evidence survived promotion into the deployment registry.

Workload security · Kubernetes / Docker / Sigstore · By Cloud Security Desk
Technical guideSource-based analysis

Verify the CloudTrail files behind an incident timeline

Digest delivery and successful validation are different evidence states. Preserve the files, metadata and validation result needed to distinguish them.

Detection & response · AWS · By Cloud Security Desk
Research noteIllustrative analysis

Threat-model the system around the model

The model endpoint is one component. The consequential paths often run through retrieval stores, orchestration identities, evaluation data, and operator tools.

AI systems · AWS / Azure / Google Cloud · By Umair Akbar and Ahmed Elshekh