Validate core file changes in an isolated Clone-01 before applying them.
npx skills add https://github.com/zkksdk/skills --skill safe-self-mod
Please help me install this Agent Skill. Skill: safe-self-mod Source: https://github.com/zkksdk/skills/tree/main/skills/safe-self-mod Command: npx skills add https://github.com/zkksdk/skills --skill safe-self-mod
This workflow ensures that changes affecting the main instance (core files under /workspace, including SOUL.md, IDENTITY.md, AGENTS.md, TOOLS.md, MEMORY.md, and configuration or skill files such as openclaw.json and skills/*) are validated in an isolated Clone-01 environment before applying to the main workspace.
Initiate a safe-mod workflow by validating a proposed change in Clone-01 and report the verification result.
High-intent search queries and answers about installing and using this skill.
To safely validate self-modifications, changes to core identity files are first tested in an isolated Clone-01 environment. This workflow automates risk checks and requires explicit user consent before applying any validated modifications to the main workspace.
An isolated verification workflow for self-modification is a three-stage process: detecting relevant file changes, validating them in a Clone-01 session, and applying them only after user confirmation. It prevents unverified changes from impacting the live workspace.
You can automate change management for configuration and skill files by triggering a safe-mod workflow that recognizes write operations. It performs isolated verification and reports PASS or FAIL outcomes, ensuring user consent prior to modification.
Yes, the self-modification safety workflow requires explicit user confirmation before applying changes. It analyzes verification outcomes from the Clone-01 session and reports a PASS with risk assessment or FAIL with issues, enabling informed consent.
Yes, you can use this isolated verification process for updates to openclaw.json and skills directories. The workflow recognizes write operations that touch configuration or skill files and validates them in Clone-01 before applying changes.
If a verification task fails during the self-modification workflow, the system reports a FAIL outcome with specific issues. Changes are blocked from applying to the main workspace until the issues are resolved and a subsequent verification passes.
oabdelmaksoudConfigure autonomous agents with IDENTITY.md, SOUL.md, and openclaw.json.
phj128Assess final quality gates and write PASS/FAIL to subtask records.
fredericpoitevinDefine and deploy Cursor hooks to customize agent-event behavior and automation via hooks.json and scripts.
yuhle-qugOptimize GitHub Actions workflows with security hardening and automation patterns.
