Tag
8 essays filed under Security.
KV caches aren’t harmless memory. Use provenance-aware eviction and metering to prevent corrupted reasoning, leakage, and runaway spend—with concrete designs.
Read essayApp verification will splinter agent skill distribution. Build signed skill bundles, offline installs, vetting APIs, and fallback runtimes to keep control.
Read essayA case for a neutral standard for cryptographically signed agent skills and runtime attestation that keeps marketplaces open while enforcing safety and provenan
Read essayA layered model—provenance, attestation, confinement, automated red teaming—against persistent-control attacks in agentic AI, with deployable steps.
Read essayHow to build agent memory you can trust: cryptographic provenance, scoped capability tokens, and SLOs for enterprise-ready, auditable recall.
Read essayHow to ship agentic AI under tightening app store rules: architect attested modules, supply-chain proofs, and hybrid delivery to keep control and security.
Read essayPublic APIs need agent-aware design: versioning for agents, capability-bound tokens, semantic rate limits, and verifiable intent contracts to curb abuse and cos
Read essayRequire cryptographic action tokens and reproducible execution transcripts for every agent effect to enforce least-privilege, auditability, and rollback.
Read essay