Input
OpenAPI JSON/YAML, MCP tools JSON, or generated agentready.json plus simulation scenarios. The current product is static and browser-first.
Public documentation
Make APIs and MCP tools ready before AI agents use them.
TimeProofs AgentReady checks tool definitions before agent exposure. It produces an AgentReady Score, findings, a human-readable report, a machine-readable agentready.json contract, and static scenario simulation outputs.
Scan OpenAPIScan MCP toolsRun simulationRead the public methodView examples
Developer documentation paths
Start with the browser scanner, reproduce locally with the published CLI, then pin the public GitHub Action in CI. No account, upload, backend, token or Pro purchase is required for Community usage.
CLI referenceInstall with
@alpha, run local scans and read exit codes.
GitHub ActionUse immutable tag or full-SHA pinning with contents: read.
Adoption guideMove from first browser scan to repository CI.
agentready.jsonInterpret score, status, rule codes and policy output.
ContributingUse current GitHub paths without unsupported process promises.
TroubleshootingSeparate invalid input, policy failures and environment issues.
Three static paths
| Path | Use it when | Input | Output |
|---|---|---|---|
| OpenAPI Scan | You have a REST API spec and want to know if the API is clear and bounded enough for AI agents. | openapi.json, openapi.yaml, openapi.yml | Score, endpoint findings, report, agentready.json |
| MCP Scan | You have MCP tool definitions and want to know if the tools are safe and clear enough to expose to agents. | mcp-tools.json with server + tools[] | Score, tool findings, report, agentready.json |
| Static Simulation | You already generated agentready.json and want to test scenario-level agent failure modes without executing tools. | agentready.json + scenario JSON | Counts, scenario results, agentready-simulation.json |
Analysis
Local checks for agent-facing clarity, dangerous actions, bounded parameters, missing limits, sensitive data, weak error guidance, and scenario-level failure modes.
Output
AgentReady Score, findings, recommendations, Markdown report, printable reports, agentready.json, and agentready-simulation.json.
Core product flow
OpenAPI / MCP / tool schema → AgentReady analysis → AgentReady Score → risk findings → recommended fixes → agentready.json → static scenario simulation → agentready-simulation.json → human report
What AgentReady checks
- Unclear operation or tool names.
- Ambiguous descriptions.
- Missing when to use guidance.
- Missing when not to use limits.
- Unbounded numeric parameters.
- Missing enums for closed fields.
- Dangerous actions without human confirmation.
- Irreversible actions.
- Non-corrective error responses.
- Sensitive data exposure and broad permissions.
What AgentReady is not
- Does not provide inline runtime request protection.
- Not a broad AI security platform.
- Not a guarantee that agents will never fail.
- Not a replacement for authorization or runtime controls.
- Not an MCP marketplace.
- Not live API or MCP execution.
Score statuses
85-100 : AgentReady 70-84 : Minor fixes 50-69 : Needs fixes 0-49 : Not AgentReady
agentready.json v0.1
agentready.json is the machine-readable contract generated from a scan. It gives CI, release review and integrations a structured view of tool risks and execution boundaries. Read the complete public specification.
{
"agentready_version": "0.1",
"source_type": "openapi",
"source_name": "example-openapi.json",
"score": 72,
"status": "Minor fixes",
"risk_counts": {
"critical": 0,
"high": 0,
"medium": 2,
"low": 0
},
"tools": []
}
agentready-simulation.json
agentready-simulation.json is generated from agentready.json plus scenario JSON. It reports whether a scenario would pass, warn, fail, or be not applicable under static rules.
{
"simulation_version": "0.1",
"scenarios_total": 1,
"pass": 0,
"warning": 0,
"fail": 1,
"not_applicable": 0,
"results": []
}
Privacy and safety
- The scanners run locally in the browser.
- Submitted API endpoints are not called.
- MCP tools are not executed.
- Static simulation does not call LLMs.
- No backend is required for the current static scanners.
- The tool identifies structural risks, not absolute safety.
TimeProofs AgentReady does not guarantee that an AI agent will never fail. It identifies structural risks that may cause AI agents to misuse APIs, tools or MCP servers.
AgentReady at a glance
The documentation hub connects browser scanners, the npm CLI, GitHub Action, reports, examples, adoption and troubleshooting. AgentReady is a product and candidate open standard for static pre-deployment analysis. It is not a runtime firewall, independent certification, official standards-body standard or guaranteed-safety system.
- Availability
- AgentReady Community is free and available. AgentReady Pro is planned and not purchasable.
- Language
- English is the current canonical language. No translated alternate route is published for this page.