🛡️

AI EngineeringTypeScript

Agent Persistence

Agent Persistence is an AI Engineering skill for Claude Code, published by jmagly in aiwg.

209 stars32 forkson jmagly/aiwgAdded 2026/09/04+1% in starsRepository updated 2026/09/03

View source on GitHub Download repo ZIP

agentic-codingai-frameworkanthropicautonomous-agentsclaude-codecodexcopilotcursordeveloper-toolsfine-tuningmulti-agentmulti-platformorchestrationprompt-engineeringsdlctraining-datawarpworkflow-automation

Install in seconds

Install Agent Persistence

Copy Agent Persistence into your Claude Code skills folder. Run the command in your terminal, or review the source on GitHub before installing.

terminal

npx degit https://github.com/jmagly/aiwg/tree/main/agentic/code/addons/agent-persistence/skills/agent-persistence ~/.claude/skills/agent-persistence

Requires Node.js. Downloads this skill only — not the rest of the repository — into your Claude Code skills folder.

Without Node.js

git clone https://github.com/jmagly/aiwg.git

Clones the whole repository, then copy the skill’s own directory into your skills folder yourself.

In this catalog

Source file: agentic/code/addons/agent-persistence/skills/agent-persistence/SKILL.md in jmagly/aiwg
Installs to: ~/.claude/skills/agent-persistence
Collection: One of 25 skills cataloged from this repository
Category: AI Engineering — 2970 skills

What Agent Persistence does

Agent Persistence defines reusable human-in-the-loop gates for persistent agent loops. Use it when you need authorization for destructive actions, policy overrides, or recovery escalation.

Agent Persistence is cataloged under AI Engineering on DirSkills. Agent Persistence comes from a repository tagged agentic-coding, ai-framework, anthropic, autonomous-agents and claude-code.

Documentation

README

Excerpt

Agent Persistence

Use this driver to activate or understand the agent-persistence addon:

aiwg use agent-persistence

The addon supplies deterministic gate definitions, not a replacement loop runtime. Compose it with agent-loop or another orchestrator that supports human checkpoints.

Canonical gates

destructive-action-gate.yaml requires authorization before destructive work.
false-positive-override-gate.yaml records reviewed policy overrides.
recovery-escalation-gate.yaml escalates bounded recovery failures.

Read the gate contract before integrating it, preserve its authorization boundary, and use the owning loop's status surface to verify activation.

This is the opening of the README. Read the full README on GitHub.

Frequently asked about Agent Persistence

What else does jmagly publish alongside Agent Persistence?
Agent Persistence is one of 25 skills that DirSkills catalogs from jmagly/aiwg, the repository it ships in. Its siblings there include Agent Loop, Agent Loop and Agent Loop External. Each one is a separate skill with its own page in this directory, installs the same way Agent Persistence does, and is maintained by jmagly in that same repository. The rest of the collection is listed on the jmagly/aiwg page.
How does Agent Persistence compare to other AI Engineering skills?
Agent Persistence ranks #2707 by stars among the 2970 AI Engineering skills in this catalog. The most-starred ones next to it are Architecture Decision Records, AI-First Engineering and Agentic OS. DirSkills ranks by the star count of the repository each skill ships in, so that order reflects how popular those repositories are rather than any review of Agent Persistence against them. Open each page to compare what they document and how they install.

More from jmagly/aiwg

Agent Persistence is one of 25 skills cataloged on DirSkills from jmagly/aiwg.

See all 25 skills →

Agent Loop

Agent Loop detects requests for iterative autonomous task execution and routes them to the right loop executor. Use it when a user wants to keep trying, fix until green, or run an explicit agent loop.

Agent Loop

Agent Loop executes a task in repeated cycles until verifiable completion criteria are met. Use it when you want an agent to keep trying, learn from failures, and report the final status.

Agent Loop External

Agent Loop External launches a long-running agent loop outside the current session with saved state. Use it for crash recovery, parallel runs, and CI/CD-backed tasks that need resumable execution.

Al External

Al External launches long-running iterative loops outside the AI session with saved state. Use it for tasks that need crash recovery, parallel runs, or CI/CD integration.

Al Memory

Al Memory lets you list, query, and clear semantic memory entries from loop iterations. Use it to review lessons learned, failure patterns, and success patterns across runs.

Auto Test Execution

Auto Test Execution runs relevant tests after code-generating agents modify source files. It is used to enforce an execute-before-return workflow and retry when tests fail.

More AI Engineering skills

Agent Persistence is one of 2970 AI Engineering skills cataloged on DirSkills, ranked #2707 by stars among them.

Browse all AI Engineering skills →

Architecture Decision Records

Architecture Decision Records captures architectural decisions made during Claude Code sessions as structured ADRs, recording context, alternatives, and rationale. Use it when significant technical choices are made so future developers understand why the codebase is shaped the way it is.

AI-First Engineering

AI-First Engineering defines process, review, and architecture standards for teams where AI agents produce most implementation output. Use it to set ownership rules, quality gates, and evaluation criteria for agent-generated code.

Agentic OS

Agentic OS codifies a persistent multi-agent operating system architecture for Claude Code. Use it when building a multi-agent workflow with file-based memory, slash commands, and scheduled automation that survives session restarts.