AUTONOMOUS_CONTINUATION_PLAN

An implementation plan for keeping long-running NodeBench research jobs moving after agent errors or infrastructure failures. It combines rollback and lessons for semantic mistakes with model failover and budget checks for service failures such as rate limits.

In plain words
What is it for?
Use it to design self-healing runs, rollback behaviour, failure lessons, capability-aware model switching, rate-limit handling, and spending safeguards.
Why use it?
It addresses runs that stop mid-task because an agent keeps making a broken change or because a model request fails, including when the user is away.

Agent

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/homenshum/nodebenchai/autonomous_continuation_plan
Clone the repo
git clone --depth 1 https://github.com/HomenShum/NodeBenchAI
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 5,534 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.05534
Opus 5 $0.00000 $0.02767
Sonnet 5 $0.00000 $0.01107
Haiku 4.5 $0.00000 $0.00553

Measured 2d ago against content hash 447da23b2d85, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

AUTONOMOUS_CONTINUATION_PLAN scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

docs/agents/AUTONOMOUS_CONTINUATION_PLAN.md · 434 lines

How it starts

The opening of the file, as written. The whole thing — 434 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Autonomous Continuation System — Implementation Plan

A unified NodeBench subsystem enabling agents to self-heal from both semantic errors (rollback + lessons) and infrastructure failures (capability-aware model failover + budget gates), so long-running research jobs keep making progress when the user steps away.


Continuation Context

This plan picks up after the cockpit-parity work landed via:

  • PR #92feat(chat): full ChatStream port with streaming, attachments, tools
  • PR #93feat(topnav): kit avatar status panel (pulse + watching + plan + sessions + theme + links)
  • PR #105fix(theme): dark mode token overrides for kit-scoped surfaces
  • PR #115fix(avatar): wire dead Upgrade to Team button

Production cockpit is now in exact kit parity for Home, Reports, Chat, Inbox, Me, and the avatar status panel. The next failure mode that blocks dogfood sessions is agent runs that die mid-task — either from semantic spirals (the agent breaks a file, then keeps trying to fix it) or infrastructure failures (rate-limit 429s with no failover). This plan addresses both.


Scope Decision: What Applies to NodeBench

Subsystem from upstream spec Applies? Why
Visual refinement (gpt-image-2 → wireframes) No (defer indefinitely) NodeBench is a research/entity-graph platform, not a design tool. No wireframe artifacts, no Electron main process, no ui_kits/ concept.
Report refinement (NodeBench analog) Defer to v2 Conceptually: "premium pass on a research report preserving all entities/claims/citations." Nice-to-have, not core to the continuation problem.
Time-travel + learning (rollback + lessons) Yes (core) Universal agent-spiral failure mode. Convex threads, action timeline, and schema already support most plumbing.
Auto-routing + failover Yes (core, partially shipped) autonomousModelResolver.ts already has retry/jitter for 429s. Missing: capability tiers, visible switches, budget gates, cross-provider failover.

Read the full file on GitHub · 434 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 434 lines · 0 tokens per session scan A 447da23b2d85

Subscribe to this mod's changes

AUTONOMOUS_CONTINUATION_PLAN is an agent published in the GitHub repository HomenShum/NodeBenchAI (14 stars, last pushed 19d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 5,534 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other agents, from other repositories

ai-readiness-reporter

Runs the AgentRC readiness assessment on the current repository and produces a self-contained, static HTML dashboard at reports/index.html. Explains every readiness pillar, the maturity level, and an actionable remediation plan, framed by AgentRC measure → generate → maintain loop. Use when asked to assess, audit…

github/awesome-copilot · 80 tokens

Custom Agent Foundry

Expert at designing and creating VS Code custom agents with optimal configurations.

github/awesome-copilot · 17 tokens

Agent Governance Reviewer

AI agent governance expert that reviews code for safety issues, missing governance controls, and helps implement policy enforcement, trust scoring, and audit trails in agent systems.

github/awesome-copilot · 35 tokens

chat-agents

Build AI-powered chat interfaces with AIChatAgent and useAgentChat. Messages are automatically persisted to SQLite, streams resume on disconnect, and tool calls work across server and client.

cloudflare/agents · 0 tokens

callable-methods

Callable methods let clients invoke agent methods over WebSocket using RPC (Remote Procedure Call). Mark methods with @callable() to expose them to external clients like browsers, mobile apps, or other services.

cloudflare/agents · 0 tokens

webhooks

Receive webhook events from external services and route them to dedicated agent instances. Each webhook source (repository, customer, device) can have its own agent with isolated state, persistent storage, and real-time client connections.

cloudflare/agents · 0 tokens