AIops-tools/Endpoint-AIops

> **Disclaimer**: Community-maintained open-source project. **Not affiliated with, endorsed by, or sponsored by any endpoint-management vendor.** Product and trademark names belong to their owners. MIT licensed.
Governed AI-ops for **managed-endpoint fleets** — thin clients, VDI endpoints,
and other centrally-managed devices — with a **built-in governance harness**:
unified audit log, token/runaway budget guard, undo-token recording, and
descriptive risk tiers. Vendor-neutral: it talks to an
endpoint-management server's REST API (Bearer auth) through a configurable
**dialect** — see Dialects. Self-contained: no
dependencies beyond httpx and the MCP SDK. The test suite is mock-based; the
endpoint-management REST paths have not yet been exercised against a live
management server — see docs/VERIFICATION.md.
What it does
Two signature analyses, plus the guarded reads and writes around them:
- **Login-storm analysis** — during a "everyone logs in at 9am" incident,
detect the storm (bursts of concurrent logins in a sliding window) and rank
the endpoints/users dragging login and boot times. Every flag is reported
with its number, not a black-box verdict.
- **Patch / config drift** — find endpoints that have drifted from the fleet
(outdated patch level, stray agent version, divergent OS build or config
profile). With no declared baseline it derives one by fleet majority, so it
works before a gold image exists.
What works
- **CLI** (endpoint-aiops ...): init, overview, endpoint list/get/assign-profile/reboot, session list/storm, drift report/patch, secret set/list/rm/migrate/rotate-password, doctor, mcp.
- **MCP server** (endpoint-aiops mcp or endpoint-aiops-mcp): **13 tools** (10 read, 3 write), every one wrapped with the bundled @governed_tool harness.
- **Encrypted credentials**: the management-server API key lives in an encrypted store ~/.endpoint-aiops/secrets.enc (Fernet + scrypt) — **never plaintext on disk**. Unlock with a master password from ENDPOINT_AIOPS_MASTER_PASSWORD (MCP/CI) or an interactive prompt (CLI).
- **Reversibility**: endpoint_assign_profile (high risk) captures the prior profile and records an inverse "reassign the prior profile" undo descriptor. endpoint_reboot (medium risk) captures the prior online state for the audit record but declares no undo (a reboot has no safe inverse).
- **Safety**: state-changing CLI ops (endpoint assign-profile, endpoint reboot) require double confirmation and support --dry-run.
Capability matrix (13 MCP tools)
| Category | Tools | Count | R/W |
|----------|-------|:-----:|:---:|
| **Overview** | overview | 1 | read |
| **Inventory** | endpoint_list, endpoint_get, endpoint_health_score | 3 | read |
| **Sessions** | session_list, login_storm_analysis | 2 | read |
| **Drift** | drift_report, patch_status, patch_compliance | 3 | read |
| **Remediation** | endpoint_assign_profile | 1 | write (high) |
| | endpoint_reboot | 1 | write (medium) |
| **Undo** | undo_list | 1 | read |
| | undo_apply | 1 | write (medium) |
The analysis tools (login_storm_analysis, drift_report, patch_status,
patch_compliance, endpoint_health_score) accept injected records for
pure/offline analysis; endpoint_health_score and patch_compliance are
injected-only, the others also pull live from a configured target.
What this tool does, and does not, decide
It delivers managed-endpoint operations — reads and writes — accurately and
efficiently, and records every one of them. It does **not** decide whether a
write is allowed to happen. That is the agent's judgement, or the permission of
the account you connect it with: give it a management-console account or API
token scoped to a read-only role and the writes fail at the server — the place
that actually owns the permission.