Notifications
Clear all
Topic starter
01/08/2026 9:00 pm
TL;DR: Runtime detection, session tracing, and inline blocking for Claude Chat, Cowork, and Code are now framed around compromised Skills, MCP servers, and indirect prompt injection, according to Straiker. The deeper issue is that agent governance fails when untrusted content becomes an execution path, not just a prompt.
NHIMG editorial — based on content published by Straikerai: How Straiker Secures Claude with the Claude Compliance API
By the numbers:
- 98.1 to 99 percent accuracy, with a 0.02 percent false-positive rate and roughly 306 milliseconds of added latency.
- 36 percent of successful coding-agent attacks reached remote code execution on the developer's machine.
- 4,242 of the 17,651-plus MCP servers we track carry a known vulnerability.
Questions worth separating out
Q: How should security teams govern AI agents that can access enterprise systems?
A: Security teams should govern AI agents as non-human identities with explicit ownership, scoped privileges, and continuous monitoring.
Q: Why do AI agents create new risk in non-human identity management?
A: AI agents create risk because they operate as software identities with delegated authority, but many organisations do not track them with the same discipline applied to users or service accounts.
Q: What breaks when Skills and MCP servers are approved without full inventory and review?
A: You lose visibility into the content and tools that can silently change agent behaviour.
Practitioner guidance
- Inventory Claude-connected Skills and MCP endpoints Build and maintain a complete register of every Skill, connector, and MCP server an agent can reach.
- Enforce runtime blocking for high-risk agent actions Use inline controls for tool calls that can move data, access browsers, or trigger external systems.
- Retain session-level evidence for investigations Preserve prompts, responses, tool calls, and the exact content path that influenced the agent decision.
What's in the full article
Straiker's full blog covers the operational detail this post intentionally leaves for the source:
- Runtime detection and inline blocking details for Claude Chat, Cowork, and Code deployments
- Session reconstruction examples showing how prompts, tools, and data paths are replayed for investigations
- Skills and MCP scanner logic for classifying indirect prompt injection and data exposure risk
- Framework mapping specifics for OWASP, NIST AI RMF, MITRE ATLAS, and EU AI Act alignment
👉 Read Straiker's analysis of Claude compliance API coverage and agent security →
Claude compliance API coverage: are runtime controls enough for agents?
Explore further
01/08/2026 9:07 pm
Runtime agent security is becoming a control-plane issue, not just a model safety issue. Claude coverage across Chat, Cowork, and Code shows that agents now move through real enterprise workflows, which means the decision point is the tool call, not the chat turn. When a security control can block inline and trace sessions, it begins to look more like identity enforcement than content moderation. Practitioners should treat agent execution paths as governed access paths, not experimental AI features.
A question worth separating out:
Q: Who is accountable when an AI agent accesses sensitive data it was not meant to use?
A: Accountability sits with the team that approved the agent, its connectors, and its policy boundaries, not with the runtime behaviour alone. Organisations need ownership for intent, permissions, monitoring, and validation so they can prove whether the agent stayed inside its approved purpose. Without that, audit and regulatory response become retrospective guesswork.
👉 Read our full editorial: Claude compliance API coverage exposes the limits of runtime agent security
