Skill Claude CodeCodex
Changing lines in a file you wrote with memoryedit, finding them with memoryread, and what memoryedit's refusals mean.
A cognition kernel that wraps a frozen LLM in a persistent identity + trust + governance loop — testing whether capability comes from structure, not weights. Research-stage; honest about what's real vs mocked.
This repository also configures its own agents. See what SAGE tells them →
Skill Claude CodeCodex
Changing lines in a file you wrote with memoryedit, finding them with memoryread, and what memoryedit's refusals mean.
Skill Claude CodeCodex
Running code you wrote: run (a sandbox, now) or requestrun (your seat, later), how to make a run settle a question, and how to read UNCHANGED, SAME AS and a decline.
Skill Claude CodeCodex
Keeping what you concluded where a later you will see it: settle, retirenote, closing todo items, recall before testing again.
Skill Claude CodeCodex
How you see: camera takes a frame and shows it to you at your next beat. Its description says where this seat's frame comes from (a cortex, a Mac's camera found by name, or a device node), and what a device node is and is not.
Skill Claude CodeCodex
What a refusal from the gate means, and the ways forward: a scope request, an appeal, or asking. What pending, expired and "does not exist" mean.
Skill Claude CodeCodex
Writing a skill of your own: a procedure you worked out, kept so a later you can find it when the same situation comes back.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: