AgentsEnterprise AI Agent Security 80 Percent Report Risky BehaviorsHelp Net Security·high signalXBlueskyLinkedInCopy linkAIUC-1 Consortium: 80% enterprises observe risky agent behaviors, only 21% have complete visibility into agent permissions. 1,200 unofficial AI apps average.SourceSource pageHelp Net Security↳ Follow the threadStack layer / Threat patternElva Launches Against Postman With Flat Workspace Pricing and Specs Generated From Repo CommitsElva (surfaced via the Product Hunt daily leaderboard for 2026-09-14)Stack layer / Threat patternTip: read a built-in Claude Code plugin's behaviour by running it from source with `--plugin-dir`GitHubThreat pattern / ContrastConstellation Research: Salesforce's Agent Pricing Is Confusing Buyers, With Spend Up 30% Against ~3% Productivity GainsConstellation ResearchStack layer / Threat patternA first-hand report on Anthropic's Claude Certified Architect exam says it is not a docs-skimr/ClaudeAI (exam details via anthropic.skilljar.com)Stack layer / Threat patternMarket Concentration Barely Changes Model Collapse: Pushing One Model to 90% Share Moves Five-Generation Endpoints a Few PercentarXiv 2609.11146Stack layer / Threat patternPattern: Anthropic's own bar is that Claude-written production code gets reviewed harder than human-written codeSimon WillisonThreat pattern / ContrastA Harness Survey of 700 Enterprise Engineers Found 77% Claim a Full Agent Inventory but Only 44% Run Discovery ToolingHarness / PR NewswirePolicy dependency / Stack layerRIPPLE: an edit confined to one prompt-policy segment changes downstream behavior, so replay candidate edits after previously accepted ones before persistingarXiv 2609.12127