Nanoclaw Repl
Run and extend NanoClaw v2’s markdown-backed REPL (claw.js) with sessions, branching, compaction, and exports while keeping zero external runtime dependencies.
Overview
NanoClaw REPL is an agent skill most often used in Build (also Operate iterate) that documents how to run and extend ECC’s zero-dependency, markdown-session REPL built on claude -p.
Install
npx skills add https://github.com/affaan-m/everything-claude-code --skill nanoclaw-replWhat is this skill?
- Persistent markdown-backed sessions with `/branch`, `/search`, `/compact`, and `/export` (md/json/txt)
- Runtime controls: `/model` switching and dynamic skill loading via `/load`
- Session metrics via `/metrics` for lightweight observability of REPL usage
- Extension rules: zero external runtime dependencies and markdown-as-database compatibility
- Operating guidance: branch before high-risk changes, compact after milestones, export before sharing
- 8 documented slash capabilities (/model, /load, /branch, /search, /compact, /export, /metrics, plus session persistence)
Adoption & trust: 4.2k installs on skills.sh; 210k GitHub stars; 3/3 security scanners passed (skills.sh audits).
What problem does it solve?
Your agent work sprawls across chat threads with no branchable history, skill hot-loading, or exportable session trail tied to local markdown files.
Who is it for?
Solo developers extending Everything Claude Code’s NanoClaw REPL who want local, dependency-free session management and deterministic slash commands.
Skip if: Teams needing hosted multi-user agent orchestration, cloud session sync, or payment and browser automation outside the documented claw.js surface.
When should I use this skill?
When running or extending `scripts/claw.js`, NanoClaw v2 session-aware REPL operations, or ECC zero-dependency claude -p workflows.
What do I get? / Deliverables
You operate claw.js with stable sessions, safer branching before risky changes, compacted history, and exported md/json/txt artifacts you can share or archive.
- Exported session files (md/json/txt)
- Branched session state in markdown store
- Compaction-reduced session history
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
Build / agent-tooling is the canonical shelf because the skill documents operating and extending the local agent REPL, not shipping end-user product UI. Agent-tooling matches persistent sessions, dynamic `/load` skills, and REPL command handlers for Claude `-p` workflows.
Where it fits
Run a task-focused REPL session, `/load` a feature skill, and `/branch` before refactoring core scripts.
Extend deterministic slash-command handlers while preserving markdown-as-database storage.
`/compact` after shipping a milestone and `/export` json for a postmortem or handoff note.
`/search` prior sessions to find the decision log before merging a risky agent workflow change.
How it compares
Local REPL session skill for claw.js—not a hosted agent platform or generic terminal cheat sheet.
Common Questions / FAQ
Who is nanoclaw-repl for?
Builders using Everything Claude Code who operate or extend `scripts/claw.js` and want markdown-backed, session-aware agent REPL workflows.
When should I use nanoclaw-repl?
Use it during Build when wiring agent tooling and during Operate when maintaining long REPL sessions, branching risky experiments, compacting history, or exporting before reviews.
Is nanoclaw-repl safe to install?
It describes local REPL operations without mandating network calls in the skill text; confirm repo trust and review the Security Audits panel on this Prism page before running claw.js in sensitive directories.
SKILL.md
READMESKILL.md - Nanoclaw Repl
# NanoClaw REPL Use this skill when running or extending `scripts/claw.js`. ## Capabilities - persistent markdown-backed sessions - model switching with `/model` - dynamic skill loading with `/load` - session branching with `/branch` - cross-session search with `/search` - history compaction with `/compact` - export to md/json/txt with `/export` - session metrics with `/metrics` ## Operating Guidance 1. Keep sessions task-focused. 2. Branch before high-risk changes. 3. Compact after major milestones. 4. Export before sharing or archival. ## Extension Rules - keep zero external runtime dependencies - preserve markdown-as-database compatibility - keep command handlers deterministic and local