Overview Get Started

Codimir Codec SDK

Codimir AI Fabrick

Architecture MCP Layer Metrics & SLAs Orchestration Overview Roadmap Roles Standard Flow (Light ↔ Heavy)Whitepaper — Codimir Fabrick

Continuations Escalation QA / Proctor RunPod Offload

Codimir AI FabrickPlaybooks

Continuations

Checkpointing and resuming long tasks across serverless limits.

Continuations Playbook

Vercel functions have a hard 800s execution cap. Continuations solve this by checkpointing state and resuming.

Steps

Agent runs with a TimeBudget (soft = 720s).
Before timeout, agent saves Checkpoint to S3/R2.
Emit a ContinuationToken (HMAC-signed).
Client (or scheduler) POSTs to /api/agents/continue with token.
Agent resumes from checkpoint until completion.

sequenceDiagram
    participant Agent
    participant S3
    participant Client
    participant NextAPI

    Agent->>S3: Save checkpoint
    Agent-->>Client: SSE continue:{token}
    Client->>NextAPI: POST /continue {token}
    NextAPI->>S3: Load checkpoint
    NextAPI->>Agent: Resume execution

Thought to Task

From verbal prompts today to BMI telepathic execution tomorrow.

Escalation

When and how to hand off from light to heavy agents or to other roles.

On this page

Continuations Playbook Steps

AI Assistant

Assistant

Hi! I'm your AI assistant. Ask me anything or describe your task.