| Tool | Purpose | How to Use |
|---|---|---|
| 37-data-model | Single source of truth for all project entities | GET http://localhost:8010/model/projects/50-eva-ops |
| 29-foundry | Agentic capabilities (search, RAG, eval, observability) | C:\eva-foundry\eva-foundation\29-foundry |
| 48-eva-veritas | Trust score and coverage audit | MCP tool: audit_repo / get_trust_score |
| 07-foundation-layer | Copilot instructions primer + governance templates | MCP tool: apply_primer / audit_project |
Agent rule: Query the data model API before reading source files.
Invoke-RestMethod "http://localhost:8010/model/agent-guide" # complete protocol
Invoke-RestMethod "http://localhost:8010/model/agent-summary" # all layer countsOperations-level tooling to monitor, manage and remediate agent fleets across Local, GitHub and Azure planes. Provides a single admin surface (EVA Admin Console) plus agentic services and scheduled cloud watchers to keep agent runtimes healthy when developer workstations are offline.
- Consolidate watchdog scripts and GitHub workflow created in
29-foundry/tools/. - Provide an operations web console to: view agents, start/stop jobs, kill rogue processes, and review execution logs.
- Define the execution data model required to record runs, executions, actions, and evidence receipts.
- Provide automated scheduled remediation (GitHub Actions cron + Azure runbook integration).
Local fixes (close VS Code) help when you are at your workstation. When systems are running in the cloud, operations tooling is required so agents cannot go rogue while everyone's asleep. This project formalizes the processes, agents, UI and data model that make that possible.
tools/: operational scripts (watchdogs, cleaners, deploy helpers)server/: FastAPI admin API and agent orchestratorui/: Admin Console (React) skeleton and pagesmigrations/: data model migration scripts (if we extend the central model)docs/: design and runbooks
See PLAN.md for concrete tasks to bootstrap the project.
Phase 2 tasks T01-T10 are defined and ready to execute.
See docs/20260226-design.md for the full EVA relevance assessment, data model
layers (L27 executions / L28 run_actions), API spec, screen list, enhancements,
and inter-project dependency map.
- Install the workspace venv: use
C:\eva-foundry\.venv\Scripts\python.exe. Do not use system Python. - Run local health check:
powershell -File C:\eva-foundry\29-foundry\tools\watch-eva-all.ps1 - Review
29-foundry/.vscode/mcp.jsonto ensure unready MCP servers remain disabled.
Ops: marco.presta