The Test Runner Tool
Execute a project’s tests in the sandbox and return which passed, which failed, and why — so an agent can verify a change actually works before proposing to merge it.
A suggestion isn’t a shipped change
An agent that can only propose code still leaves all the work to a human. To actually deliver, it needs a real, governed workspace where it can run code, edit files, test, and use Git — safely, and without touching anything you didn’t allow.
Read-only agents
Suggestions still require a human to run, test, and commit everything.
Unsafe execution
Running agent-generated code on real infrastructure is a security risk.
No verification
Without tests and builds, an agent can’t know its change works.
Ungoverned Git
Direct repo access with no policy or audit is a non-starter in the enterprise.
Test Runner, without the risk
Capability
What it does
Run the test suite and report pass/fail.
it runs the project’s test suite in the sandbox and returns results, failures, and output.
Assignable to any agent
How it works
Predictable, inspectable behavior
Designed to be reliable.
it detects or accepts the test command, runs it under resource limits, and parses the results, so an agent gets a definitive pass/fail signal rather than a guess about whether code works.
Every call logged
Governance
Private, governed, on-premise
Runs inside your perimeter.
Execution runs in an isolated, on-premise sandbox scoped per tenant with full command and file audit logging, so an agent can do real work on your code without unsafe access or anything leaving your perimeter.
Per-tenant, logged
Parameters
The test_runner tool accepts these inputs when an agent calls it. Required inputs are flagged.
default: . Optional Project directory.
Where Test Runner pays back
Change verification
Prove a fix works before opening a PR.
Regression guard
Catch breakage an agent’s edit introduced.
TDD loops
Let an agent iterate until tests pass.
Gate merges
Block a change that fails the suite.
Assigned to agents, orchestrated as networks
On VDF AI, an industry’s use cases map to agents, and you assign tools like this one to those agents. Compose multiple agents into a governed, on-premise network.
What changes after you assign it
Questions about the Test Runner tool
What is the Test Runner tool?
It runs the project’s test suite in the sandbox and returns results, failures, and output. Assigned to a VDF AI agent, it runs under role-based policy with full audit logging so the capability is safe to use in production.
Does it detect the test command?
Yes, for common stacks; you can also pass an explicit command or a filter to run a subset.
What does it return on failure?
The failing tests and their output, so an agent can diagnose and fix before retrying.
Does it run on-premise?
Yes. Like every VDF AI tool, it can run on-premise or in your sovereign cloud, scoped per user and audit-logged, so your data never leaves your perimeter.
How do agents use it?
You assign the tool to an agent under a role-based policy; the agent calls it as one step in a task, and several agents and tools can be orchestrated together as a governed VDF AI Network.
Tools that work well alongside this one
Where this tool delivers value
Put Test Runner to work
See the Test Runner tool assigned to an agent and orchestrated in a governed, on-premise network.