Declared versus actual behavior in agent skills
ATR evaluates whether a skill’s behavior exceeds its declared purpose.
Why it matters
Tools inherit authority. A tool may operate with a user’s credentials, files, network access, or other permissions. Its capability and blast radius matter alongside code quality.
In the registry
ATR evaluated this on 210 skills and plugins in the 20 published reports, as of Nov. 9th, 2026, and found an issue in 2, across 2 reports.
Each report shows the exact file and lines, why the issue matters, and the recommended fix, with a fix prompt you can paste into your coding agent. Browse the reports.
Questions
What does ATR check for declared versus actual behavior?
ATR evaluates whether a skill’s behavior exceeds its declared purpose. Each skill or plugin is assessed against 23 rules: 14 common, one guardrail, and eight component-specific rules.
Why does declared versus actual behavior matter?
Tools inherit authority. A tool may operate with a user’s credentials, files, network access, or other permissions. Its capability and blast radius matter alongside code quality.
How often does the registry find it?
ATR evaluated this on 210 skills and plugins in the 20 published reports, as of Nov. 9th, 2026, and found an issue in 2, across 2 reports.
How do I check my own skill or plugin?
Request a scan at bluerock.io/scan. Your first scan is free. Each skill or plugin is assessed against 23 rules: 14 common, one guardrail, and eight component-specific rules.
Source: BlueRock, Component Level Security Analysis for MCP Servers, Agent Skills and Plugins (Oct. 22nd, 2026).