القائمة
Use when a repo does not yet have canonical memory, AGENTS.md, or CLAUDE.md. Bootstraps the standard Codex + Claude Code project memory layout in one step. Triggers: "project-mem-init", "bootstrap memory", "initialize project memory"
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
Use when turning recent agent-session evidence into durable skill updates. Triggers: "consolidate lessons", "update skills from sessions", "prevent repeated fixes", "recurring mistakes", "capture workflow learnings"
Use when running workloads on LUMI supercomputer, including GPU job submission, PyTorch with ROCm/AMD MI250X, container workflows, and LUMI-specific Slurm configuration. Triggers: "LUMI", "MI250X", "ROCm", "AMD GPU", "CSC", "LUMI-G"
Use when integrating W&B experiment tracking into ML training pipelines, including logging strategy, run configuration, and online/offline mode management. Triggers: "W&B", "wandb", "weights and biases", "experiment logging", "wandb.log", "wandb.init", "training dashboard"
"Use when extracting Overleaf login cookies from Chrome on macOS for the VSCode Overleaf Workshop extension. Decrypts Chrome's cookie store and outputs the session cookie in the exact format the extension expects. Triggers: "overleaf cookie", "overleaf login", "overleaf workshop cookie", "get overleaf cookie", "vscode overleaf login"
Use when preparing to launch any remote or multi-GPU ML experiment. Triggers include "preflight", "sanity check", "launch validation", "before training", "GPU check", "dataset check", "environment check"
Use when pushing changes in a repository that contains git submodules, especially when the submodule has its own commits that must not be left stranded. Triggers include "push", "git push", "submodule", "commit submodule", "workspace push"
| name | project-mem-init |
| description | Use when a repo does not yet have canonical memory, AGENTS.md, or CLAUDE.md. Bootstraps the standard Codex + Claude Code project memory layout in one step. Triggers: "project-mem-init", "bootstrap memory", "initialize project memory" |
Bootstrap shared project memory for a repo with one canonical Markdown source and generated host files.
Default layout:
memory/project.md as the canonical shared memoryAGENTS.md generated from memory/project.mdCLAUDE.md generated as a thin @AGENTS.md wrappermemory/claude.md for optional Claude-only notesREADME.md as a starter project introductionFrom the skill directory:
python scripts/project_mem_init.py --repo /path/to/repo
If the target repo already has a CLAUDE.md or AGENTS.md with real content (not a generated wrapper), that content is automatically migrated into memory/project.md before initialization. This preserves existing project instructions while adopting the canonical memory layout.
After initialization, the target repo should contain:
memory/project.md — canonical shared memory (migrated from existing files or template)memory/claude.md — Claude-specific overlayAGENTS.md — generated from memory/project.mdCLAUDE.md — generated as @AGENTS.md wrapperREADME.md — starter template (skipped if already exists)memory-sync.| Anti-Pattern | What goes wrong | Fix |
|---|---|---|
Running project-mem-init on a repo that already has memory/project.md | Overwrites canonical memory and loses history | Use memory-sync instead — project-mem-init is bootstrap only |
Hand-writing CLAUDE.md right after init | Drift on the next memory-sync sync | Edit memory/project.md; CLAUDE.md is generated |
| Skipping the auto-migration step on an existing repo | Pre-existing instructions in old CLAUDE.md/AGENTS.md get lost | Let project-mem-init migrate; review the resulting memory/project.md |
Treating project-mem-init as the ongoing sync workflow | No drift detection, no idempotence guarantee | Use memory-sync for updates after bootstrap |
| Initializing without an existing git repo | Generated files land in an un-tracked state and may be lost | git init (or run inside an existing repo) before project-mem-init |
memory-sync — Ongoing sync, drift check, and repair (use after project-mem-init)meta-init — Install tao-research-skills globally (different "init" — installs skills, not memory)agents-md-writing — Patterns for writing effective AGENTS.md/CLAUDE.md content