How agent-drivability is scored
Agent Friendliness (C2) measures how much verified programmatic access a tool exposes to a coding agent. The score is based on documented MCP, CLI, API, scripting, and file-format surfaces, not on marketing copy.
This page keeps the rubric data on-site. Longer-form workflow analysis and opinionated AI commentary live on Building Probable.
C2 is one of the five rubric criteria explained in how we grade. Higher rungs mean a tool can be driven directly from an agent or script. Lower rungs mean the workflow still depends on in-app automation or manual UI work.
Access ladder
Subscribe to Building Probable
Opinionated AI workflow analysis and the longer-form companion essay live on Building Probable.
Subscribe to Building Probable →Agent-drivable tools in the catalog
111 tools currently meet this page's inclusion rule: C2 ≥ 4, or a documented MCP server, or a documented CLI.
The list is derived from each tool's verified agentAccess data and grouped by category.
Win & Set Up the Work
Design the Building
Engineer & Analyze
Produce & Issue Documents
Coordinate & Collaborate
Administer Construction
Operate the Building
Run the Practice
C2 measures documented access surfaces, not vendor claims about AI quality. For the full rubric and evidence model, see how we grade.