Skip to content

great_cto

v2.46.0 Feature

This release adds 3 notable features for engineering teams evaluating rollout.

Published 1mo AI Agents & Assistants
✓ No known CVEs patched
Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Topics

agentic-coding claude-code-plugin claude-code-skills claude-code-subagents code-review cto
+2 more
multi-agent sdlc

Affected surfaces

auth rbac

Summary

AI summary

Autopilots now require human approval via an inbox before executing irreversible writes.

Changes in this release

Feature Medium

Adds durable run store persisting across restarts in ~/.great_cto/autopilot-runs.

Adds durable run store persisting across restarts in ~/.great_cto/autopilot-runs.

Source: llm_adapter@2026-06-07

Confidence: high

 —
Feature Medium

Adds resumable runtime with startAt and approvedGates preserving safety invariant.

Adds resumable runtime with startAt and approvedGates preserving safety invariant.

Source: llm_adapter@2026-06-07

Confidence: high

 —
Feature Medium

Adds CLI commands (start/inbox/show/approve/reject/runs) and admin board /autopilot.html with shared store.

Adds CLI commands (start/inbox/show/approve/reject/runs) and admin board /autopilot.html with shared store.

Source: llm_adapter@2026-06-07

Confidence: high

 —
Feature Medium

Adds multi‑gate requirement: tax now needs two signatures before IRS e‑file.

Adds multi‑gate requirement: tax now needs two signatures before IRS e‑file.

Source: llm_adapter@2026-06-07

Confidence: high

 —
Feature Medium

Adds push notification to signer when a gate opens (🛂 new case awaiting your signature).

Adds push notification to signer when a gate opens (🛂 new case awaiting your signature).

Source: llm_adapter@2026-06-07

Confidence: high

 —
Feature Medium

Adds hardening: multi‑tenant scoping and per‑run idempotency key to prevent double submits.

Adds hardening: multi‑tenant scoping and per‑run idempotency key to prevent double submits.

Source: llm_adapter@2026-06-07

Confidence: low

 —
Feature Low

Adds 261 library tests for Autopilot functionality.

Adds 261 library tests for Autopilot functionality.

Source: llm_adapter@2026-06-07

Confidence: high

 —
Full changelog

Autopilots become a real operated service: a durable run pauses at the human checkpoint, waits in an inbox for a named licensed human to sign, then resumes and executes the irreversible action (the write). Built on rcm; works for all 16.

  • Durable run store — runs persist (~/.great_cto/autopilot-runs), survive restarts, immutable audit trail. startRun → awaiting-approval; approve(id, who) resumes + runs the write; reject ends it with nothing irreversible run.
  • Resumable runtime — runFlow gains startAt + approvedGates; the v2.43.0 safety invariant holds end-to-end (the write fires ONLY because a human signed its gate — across a process boundary).
  • CLI autopilot.mjs (start/inbox/show/approve/reject/runs) ↔ admin board /autopilot.html + /api/autopilot/* — one shared store, fully synced.
  • Multi-gate — tax now needs two signatures (preparer PTIN → taxpayer Form 8879) before the IRS e-file.
  • Push to the signer — board pushes '🛂 new case awaiting your signature' when a gate opens.
  • Hardening — multi-tenant scoping (each tenant sees only its queue) + a stable per-run idempotency key so a retried write never double-submits.
  • 261 lib tests. Proven on rcm live: 3 live connectors → pause → coder signs → 837 submitted → completed.
View diff on GitHub

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Share on X Share on Bluesky

Track great_cto

Get notified when new releases ship.

Sign up free

About great_cto

Engineering-management layer of 34 specialist AI agents covering the full SDLC (architect, PM, senior-dev, reviewer, QA, security, devops, L3-support + 18 archetype-specific reviewers) with auto-detected archetypes and compliance gates (PCI-DSS, HIPAA, FedRAMP, GDPR, EU AI Act). Runs in Claude Code, Cursor, Codex CLI, Aider, and Continue via AGENTS.md + MCP. MIT.

All releases →

Related context

Earlier breaking changes

  • v2.72.2 Removes Autopilot sidebar item and Operate topbar button from dev board.
  • v2.55.0 Hard-gates the Build board; redirects invite sessions to Operate for operators.
  • v2.43.0 Runtime now blocks autonomous execution of irreversible actions without prior human checkpoint.
  • v2.43.0 Adds reversible and blastRadius fields to every flow step.
  • v2.32.0 Removed AgentShield scanner and its CLI commands.

Beta — feedback welcome: [email protected]