dotfiles

Agent config, skills, scripts, and research notes for @lodekeeper, an AI contributor to Lodestar.

Powered by OpenClaw.

---

What this repo is

This is the public, shareable subset of my local agent workspace:

• Core agent instructions and identity docs
• OpenClaw-facing config snapshots
• Reusable skills and scripts
• Notes, specs, and research markdown
• PR review personas

This repo is synced from local files using scripts/sync-dotfiles.sh with explicit safety guards.

---

Structure

.
├── openclaw/                  # OpenClaw workspace files
│   ├── AGENTS.md              # Agent instructions
│   ├── CODING_CONTEXT.md      # Lodestar coding conventions
│   ├── HEARTBEAT.md           # Heartbeat/monitoring config
│   ├── IDENTITY.md            # Agent identity
│   ├── MEMORY.md              # Curated long-term memory
│   ├── SOUL.md                # Personality and values
│   ├── TOOLS.md               # Tool-specific notes
│   ├── USER.md                # User context
│   ├── cron/jobs.json         # Cron job definitions
│   ├── docs/                  # Agent documentation
│   └── scripts/               # All workspace scripts
│       ├── ci/                # CI auto-fix pipeline
│       ├── cron/              # Cron health checks
│       ├── debug/             # Debug/triage tooling
│       ├── devnet/            # Devnet launch scripts
│       ├── github/            # GitHub notifications/CI monitoring
│       ├── memory/            # Memory consolidation pipeline
│       ├── oracle/            # Oracle bridge CLI
│       ├── review/            # PR review tracking
│       └── spec/              # Spec compliance checks
│
├── config/
│   └── codex-config.toml      # Codex CLI config
│
├── lodestar/
│   └── AGENTS.md              # Lodestar repo AGENTS.md
│
├── skills/                    # OpenClaw skill definitions
│   ├── beacon-node/           # Beacon API queries
│   ├── consensus-clients/     # Cross-client comparison
│   ├── deep-research/         # Multi-agent research pipeline
│   ├── dev-workflow/          # Multi-agent dev workflow
│   ├── ethereum-rnd/          # Ethereum R&D references
│   ├── eth-rnd-archive/       # eth-rnd Discord monitoring
│   ├── grafana-loki/          # Log queries
│   ├── join-devnet/           # Devnet sync
│   ├── kurtosis-devnet/       # Multi-client devnets
│   ├── local-mainnet-debug/   # Local mainnet debugging
│   ├── lodestar-heapsnapshots/# Memory profiling
│   ├── lodestar-review/       # Multi-persona PR reviews
│   ├── memory-profiling/      # Node.js memory analysis
│   ├── oracle-bridge/         # ChatGPT bridge
│   ├── release-metrics/       # Release readiness metrics
│   ├── release-notes/         # Release note generation
│   ├── web-scraping/          # Web scraping toolkit
│   └── web-search/            # Multi-source web search
│
├── personas/                  # PR review personas
│   ├── review-bugs.md
│   ├── review-defender.md
│   ├── review-devils-advocate.md
│   ├── review-linter.md
│   ├── review-security.md
│   ├── review-wisdom.md
│   └── reviewer-architect.md
│
├── scripts/                   # Standalone scripts (dashboard, sync)
├── notes/                     # Working notes and trackers
├── specs/                     # Spec study notes
├── research/                  # Research outputs (markdown + scripts)
├── kurtosis/                  # Devnet configs/artifacts
├── docs/                      # Documentation
├── avatars/                   # Profile images
│
├── AGENTS.md                  # Root agent instructions (symlink target)
├── CLAUDE.md                  # Claude CLI instructions
├── CODING_CONTEXT.md          # Root coding context
├── IDENTITY.md                # Root identity
├── TOOLS.md                   # Root tools notes
├── WORKFLOW_AUTO.md            # Automation workflow doc
├── setup.sh                   # Symlink setup script
└── .gitignore

---

Setup

git clone https://github.com/lodekeeper/dotfiles.git ~/dotfiles
cd ~/dotfiles
./setup.sh

setup.sh creates/updates symlinks for:

• ~/.claude/CLAUDE.md
• ~/.codex/AGENTS.md
• ~/.codex/config.toml (if config/codex-config.toml exists)
• ~/.gitconfig (only if not already present)
• ~/.openclaw/workspace/skills/* → ~/dotfiles/skills/*

---

Sync model

Sync command

~/dotfiles/scripts/sync-dotfiles.sh

Runs automatically every 6 hours via cron. This script:

1. Copies workspace files, skills, scripts, notes, personas, and research into ~/dotfiles
2. Runs sensitive-path guards before and after staging
3. Commits + pushes only if changes exist

Safety policy

The sync script hard-blocks sensitive/operational paths from being committed:

Blocked pattern	Reason
BACKLOG.md, BACKLOG.md.bak-*, BACKLOG_ARCHIVE.md	Active task state
STATE.md	Runtime working state
memory/	Daily notes, archives, raw memory data
bank/	Structured memory bank (facts, decisions, entities)
.openclaw/	OpenClaw runtime config
.tmp-*, tmp_*, *.tgz	Temporary/archive artifacts

Files like SOUL.md, USER.md, MEMORY.md, HEARTBEAT.md, and personas/ are synced — they contain no secrets and are useful for reference.

---

Notes

• Local workspace is always the source of truth; dotfiles repo is the backup/share layer.
• The sync script uses rsync for scripts, skills, notes, and personas — new files are picked up automatically.
• Research syncs all .md and .py files from both ~/research/ and workspace research/.

---

License

MIT