{"id":45656566,"url":"https://github.com/rune-kit/rune","last_synced_at":"2026-05-01T10:04:42.400Z","repository":{"id":340292319,"uuid":"1165384120","full_name":"Rune-kit/rune","owner":"Rune-kit","description":"Less skills. Deeper connections. A lean, interconnected skill ecosystem for Claude Code.","archived":false,"fork":false,"pushed_at":"2026-04-01T05:02:01.000Z","size":1864,"stargazers_count":46,"open_issues_count":0,"forks_count":12,"subscribers_count":0,"default_branch":"master","last_synced_at":"2026-04-01T05:50:56.098Z","etag":null,"topics":["claude-code","mesh","orchestration","plugin","skills","tdd","workflow"],"latest_commit_sha":null,"homepage":null,"language":"JavaScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/Rune-kit.png","metadata":{"files":{"readme":"README.md","changelog":"CHANGELOG.md","contributing":"docs/CONTRIBUTING-L4.md","funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":"ROADMAP.md","authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-02-24T05:30:26.000Z","updated_at":"2026-04-01T05:02:05.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/Rune-kit/rune","commit_stats":null,"previous_names":["nhadaututtheky/rune"],"tags_count":16,"template":false,"template_full_name":null,"purl":"pkg:github/Rune-kit/rune","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Rune-kit%2Frune","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Rune-kit%2Frune/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Rune-kit%2Frune/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Rune-kit%2Frune/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Rune-kit","download_url":"https://codeload.github.com/Rune-kit/rune/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Rune-kit%2Frune/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":31316094,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-02T12:59:32.332Z","status":"ssl_error","status_checked_at":"2026-04-02T12:54:48.875Z","response_time":89,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.5:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["claude-code","mesh","orchestration","plugin","skills","tdd","workflow"],"created_at":"2026-02-24T08:43:41.955Z","updated_at":"2026-05-01T10:04:42.390Z","avatar_url":"https://github.com/Rune-kit.png","language":"JavaScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"\u003cp align=\"center\"\u003e\n  \u003cimg src=\"assets/banner.svg\" alt=\"Rune — Skill Mesh for AI Coding Assistants\" width=\"100%\"\u003e\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003cstrong\u003eLess skills. Deeper connections.\u003c/strong\u003e\u003cbr\u003e\n  A lean, interconnected skill ecosystem for AI coding assistants.\u003cbr\u003e\n  63 skills · 215+ mesh connections · 8 platforms · MIT\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003ca href=\"https://rune-kit.github.io/rune\"\u003e\u003cimg src=\"https://img.shields.io/badge/Landing_Page-rune--kit.github.io-blue?style=for-the-badge\" alt=\"Landing Page\"\u003e\u003c/a\u003e\n  \u003ca href=\"https://rune-kit.github.io/rune#pricing\"\u003e\u003cimg src=\"https://img.shields.io/badge/Pro_%2449-lifetime-blueviolet?style=for-the-badge\" alt=\"Rune Pro $49\"\u003e\u003c/a\u003e\n  \u003ca href=\"https://rune-kit.github.io/rune#pricing\"\u003e\u003cimg src=\"https://img.shields.io/badge/Business_%24149-lifetime-orange?style=for-the-badge\" alt=\"Rune Business $149\"\u003e\u003c/a\u003e\n  \u003ca href=\"https://t.me/xlabs_updates\"\u003e\u003cimg src=\"https://img.shields.io/badge/Telegram-Updates-26A5E4?style=for-the-badge\u0026logo=telegram\u0026logoColor=white\" alt=\"Telegram Updates\"\u003e\u003c/a\u003e\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003cstrong\u003eClaude Code\u003c/strong\u003e (native plugin) · \u003cstrong\u003eCursor\u003c/strong\u003e · \u003cstrong\u003eWindsurf\u003c/strong\u003e · \u003cstrong\u003eGoogle Antigravity\u003c/strong\u003e · \u003cstrong\u003eOpenAI Codex\u003c/strong\u003e · \u003cstrong\u003eOpenCode\u003c/strong\u003e · any AI IDE\n\u003c/p\u003e\n\n## Why Rune?\n\nMost skill ecosystems are either **too many isolated skills** (540+ that don't talk to each other) or **rigid pipelines** (A → B → C, if B fails everything stops).\n\nRune is a **mesh** — 63 skills with 215+ connections across a 5-layer architecture. Skills call each other bidirectionally, forming resilient workflows that adapt when things go wrong.\n\n```\nPipeline:  A → B → C → D         (B fails = stuck)\nHub-Spoke: A → HUB → C           (HUB fails = stuck)\nMesh:      A ↔ B ↔ C             (B fails = A reaches C via D→E)\n           ↕       ↕\n           D ↔ E ↔ F\n```\n\n## Benchmark: With Rune vs Without Rune\n\nWe ran 10 standardized coding tasks on Claude Code — once **without** Rune (vanilla), once **with** Rune — and measured tokens, cost, duration, and correctness.\n\n### Headline Results\n\n```\n                Without Rune    With Rune     Delta\nAvg Tokens:     541,400         454,491       ↓ 16%\nAvg Cost:       $0.69           $0.65         ↓ 6%\nAvg Duration:   2.3 min         2.1 min       ↓ 9%\nAvg Tool Calls: 14              13            ↓ 7%\nCorrectness:    9/10            9/10          =\n```\n\n### Where Rune Shines: Complex Tasks\n\n| Task | Difficulty | Tokens | Cost | Duration | Tools |\n|------|-----------|--------|------|----------|-------|\n| Refactor 450-line component | Medium | **-62%** | **-17%** | **-32%** | **-27%** |\n| Full feature (auth + API + tests) | Complex | **-36%** | **-29%** | **-31%** | **-27%** |\n| Add Zod validation | Easy | -9% | **-28%** | **-32%** | 0% |\n| Dark mode across 6 components | Hard | ~0% | +10% | -7% | -6% |\n\nRune doesn't make Claude smarter — Claude already knows how to code. Rune makes Claude **disciplined**. The more complex the task, the more discipline matters.\n\n\u003e _\"Without Rune, Claude writes code that works. With Rune, Claude writes code that lasts.\"_\n\n\u003cdetails\u003e\n\u003csummary\u003eFull 10-task breakdown\u003c/summary\u003e\n\n| # | Task | Diff | Tokens | Cost | Time | Correct |\n|---|------|------|--------|------|------|---------|\n| 1 | Zod Validation | Easy | -9% | -28% | -32% | ✅ → ✅ |\n| 2 | Fix N+1 Query | Easy | +12% | +25% | +3% | ❌ → ❌ |\n| 3 | Cursor Pagination | Med | +12% | +19% | -9% | ✅ → ✅ |\n| 4 | Security Review | Med | +13% | +32% | +3% | ✅ → ✅ |\n| 5 | Rate Limiting | Med | +12% | +5% | +5% | ✅ → ✅ |\n| 6 | Refactor Component | Med | **-62%** | **-17%** | **-32%** | ✅ → ✅ |\n| 7 | Dark Mode (6 files) | Hard | ~0% | +10% | -7% | ✅ → ✅ |\n| 8 | DB Migration | Hard | +52% | +11% | +49% | ✅ → ✅ |\n| 9 | Memory Leak Debug | Hard | +13% | +28% | -2% | ✅ → ✅ |\n| 10 | Full Auth System | Complex | **-36%** | **-29%** | **-31%** | ✅ → ✅ |\n\n_Methodology: Claude Code CLI headless mode (`claude -p --output-format json`), 10 tasks with fixture code, pattern-based correctness evaluation. Source: [`Benchmark/`](Benchmark/)_\n\n\u003c/details\u003e\n\n---\n\n## What's New (v2.16.0 — Skill Enrichment + Triage Workflow + Output Modes)\n\n- **`debug` v1.2.0 — Step 0: Build Feedback Loop** — 10-rank ladder (failing test → curl → CLI snapshot → headless browser → trace replay → throwaway harness → fuzz → bisection → differential → HITL script). Codifies \"the loop is the speed limit\" — a fast deterministic pass/fail signal turns debugging into mechanical bisection. Skip if repro is already \u003c 5s and deterministic; \u003e 10 min loop construction triggers 3-Fix Escalation (architecture is the problem).\n- **`plan` v1.6.0 — Vertical Slice Mode** — tracer-bullet task decomposition. Each task = end-to-end path through schema + API + UI + test, demoable on its own. AFK / HITL classification. Stops \"horizontal layer\" planning that blocks on the slowest layer.\n- **`context-engine` v1.2.0 — Caveman Output Mode** — auto-activates on context ORANGE / RED (or `/caveman`). Strips filler / articles / hedging / pleasantries while preserving full technical accuracy (~75% output reduction). Auto-clarity exceptions for security warnings, destructive-action confirmations, multi-step sequences, root-cause diagnosis.\n- **`ba` v0.13.0 — Synthesis Mode + Out-of-Scope WRITE** — when prior conversation has rich context (pasted spec, \u003e 1000 words, continuation session), extract Requirements Document with source citations and confirm instead of re-interviewing. Step 1.6 closes the `.out-of-scope/` write loop: explicit mid-elicitation rejections produce a durable `.out-of-scope/\u003cslug\u003e.md` record so future sessions don't re-litigate.\n- **`context-pack` v0.3.0 — Agent Brief Variant** — durable handoff format for AFK agents (issue tracker queues, autopilot multi-session work). Behavioral over procedural; type names over file:line. Survives codebase drift between handoff and execution.\n- **`review-intake` v1.3.0 — Issue Triage Mode** — new mode for issue tracker items (PR Review remains default). State machine: needs-triage → needs-info / ready-for-agent / ready-for-human / wontfix. Repro-first HARD-GATE for bugs (calls `debug` Step 0 if multi-component). Vague issues route to `ba` Synthesis Mode for grilling. AGENT-BRIEF emission for `ready-for-agent`.\n- **5 new mesh signals** — `output.density.set`, `triage.classified`, `agent.brief.ready`, `outofscope.recorded` + `EXTERNAL_TRIGGER_SIGNALS` whitelist concept (symmetric to `INTENTIONAL_BROADCAST_SIGNALS`).\n- **Validator cleanups** — `validate-skills.js` Done-When regex relaxed (scope-aware, supports mode-based subsections); 9 pre-existing validation errors cleaned. `validate-signals.js` gained `EXTERNAL_TRIGGER_SIGNALS` set.\n- **Provenance** — second graft pass from [`mattpocock/skills`](https://github.com/mattpocock/skills) (MIT). Round 1 had silently grafted 7 patterns (improve-architecture, CONTEXT.md, design-it-twice, zoom-out, oracle-mode, grill, out-of-scope); Round 2 + 2b documented + extended.\n\n### Previous (v2.15.0 — Second Opinion + Cross-Provider + Routing Clarity)\n\n- **`adversary` v0.2.0 — Mode: oracle** — when `agent.stuck` fires from `debug` (3 disproved hypotheses) or `fix` (2+ failed attempts), oracle-mode dispatches a stateless second-model pass with explicit \"no prior context\" framing. Bundle format is regex-validated (`[SYSTEM]` invariant role-priming + `[USER]` template + `### File N`), token-capped (100k bundle, 4k per file, 12 files max), citation-required reply contract. Secrets auto-redacted. Breaks the confirmation-bias loop that scout's zoom-out (structural pivot) cannot.\n- **`session-bridge` v0.8.0 — Detach Mode** — async escalation primitive. Heavy-model second-opinion calls (1-10 min wall time) no longer block the primary agent. `.rune/oracle-pending/\u003csessionId\u003e.json` is the rendezvous file; idempotent dispatch (bundleHash-keyed); 10min default timeout; 24h orphan cleanup on session start. `cook` Phase 4 and `team` Phase 3 reattach via filesystem poll between adjacent tasks.\n- **`context-engine` v1.1.0 — Mode: preview** — pre-flight cost gate. Caller emits `context.preview` BEFORE bundle build with file list + estimated tokens (chars × 0.25). Per-caller thresholds: adversary 50k/100k, team 30k/80k (per worker), review 40k/100k, audit 60k/120k. Action enum `proceed | warn | block`. Override via `RUNE_CONTEXT_THRESHOLDS_\u003cCALLER\u003e`. Stops `team` parallel workstreams from silently blowing $20 of Opus tokens.\n- **Cross-provider model mapping** — 5 non-Anthropic adapters now translate `model: opus|sonnet|haiku` to provider-correct names. **codex** → gpt-5-pro / gpt-5 / gpt-5-mini. **antigravity** → gemini-3-pro / gemini-3-flash / gemini-3-flash-lite. **opencode / openclaw / generic** → tier:heavy / tier:mid / tier:light (provider-agnostic). claude / cursor / windsurf remain no-op (Anthropic backend understands native names).\n- **Routing clarity sweep** — all 63 SKILL.md descriptions now double-quoted (YAML safety). 13 ambiguous-name skills got explicit \"Use when…\" routing hints so skill-router doesn't have to guess: ba, completion-gate, constraint-check, doc-processor, integrity-check, logic-guardian, onboard, preflight, sentinel-env, watchdog, worktree, hallucination-guard, mcp-builder.\n- **4 new mesh signals** — `oracle.dispatched`, `oracle.response`, `oracle.failed`, `context.preview`. All registered in Signal Catalog with full emit/listen mapping. `agent.stuck` listeners updated to include adversary in addition to scout.\n- **1,331 tests** — +71 from v2.14.0 across 5 new test files: adapter-model-mapping (18), oracle-bundle-format (19), oracle-pending-schema (16), context-preview-signal (13), skill-description-quality (5).\n\n### Previous (v2.14.0 — Deep Modules)\n\n- **`improve-architecture` skill (NEW L2, opus)** — controlled vocabulary (Module / Interface / Implementation / Depth / Seam / Adapter / Leverage / Locality), numeric depth-leverage-locality scoring (1–5 each), 4 dependency categories, structured proposal payloads.\n- **TDD vertical-slicing HARD-GATE** — `test` v1.3.0 catches \"horizontal slicing\" (5 tests before any GREEN), commit-pair audit trail, shape-test smell detector.\n- **`.out-of-scope/` knowledge base** — `ba` v0.11.0 reads, `review-intake` v1.2.0 writes. Stops re-litigation of rejected features.\n- **CONTEXT.md inline-sharpen + ADR 3-criteria gate** — `journal` v0.4.0 only opens an ADR when sum ≥ 11 + each axis ≥ 3.\n- **Agent Brief durability** — `context-pack` v0.2.0 regex smell tests block stale paths/line numbers.\n- **Design-It-Twice mode** — `brainstorm` v0.6.0 with constraint-pinned parallel subagents + diversity score gate.\n- **Zoom-out + explore-first micro-utilities** — `scout` v0.4.0 listens for `agent.stuck`; `ba` Step 2.0 HARD-GATE requires tool-call evidence.\n\n### Previous (v2.13.0 — Script Contract + Media Pack)\n\n- **`@rune-pro/media` pack v1.0.0** — new Pro pack: raster image generation across 5 providers (Codex CLI, DALL-E, Replicate, Stability AI, local SD), prompt engineering with 4-gate safety check (trademark, public-figure, prompt-injection, uncanny-precondition), batch asset pipeline with multi-resolution variants + WebP/AVIF conversion + EXIF strip.\n- **`sentinel-env` v0.3.0** — 9-tier binary detection for hard-dependency checks.\n- **`skill-forge` v1.8.0** — new Phase 5.25 \"Script Contract\" — helper scripts must follow stdout=paths / stderr=diagnostics / `--json` opt-in / semantic exit codes. HARD-GATE on pre-ship verification.\n- **OpenClaw adapter** — `generateManifest` now declares `artifactConvention`.\n\n### Previous (v2.12.0 — Auto-Discipline)\n\n- **Runtime auto-discipline** — `rune hooks install` wires native hooks on Claude Code, Cursor, Windsurf, Antigravity so `preflight`, `sentinel`, `completion-gate` auto-fire before tool use. No more \"remember to invoke the skill.\"\n- **Three presets** — `strict` (blocking gates), `gentle` (warnings, default), `off` (uninstall). Idempotent install / uninstall with full restore of user hooks.\n- **Multi-tier hook layering** — `--tier pro` / `--tier business` stack paid-tier hooks on top of Free using a tier-tagged manifest at `$\u003cTIER\u003e_ROOT/hooks/manifest.json`. Free compiler stays tier-agnostic (MIT-clean).\n- **logic-guardian v0.3.0** — `rune init` now auto-seeds `.rune/INVARIANTS.md` with project-detected rules (money math, state machines, payment flows). Preflight reads it as a hard gate.\n- **session-bridge v0.7.0** — emits `context.compact.imminent` signal; cook/plan/team listen and checkpoint work before compaction.\n- **autopilot v1.1.0** (Pro) — honors the hooks manifest; runs overnight with the same blocking gates your interactive sessions enforce.\n- **Security** — tier name sanitization (path-traversal-safe), precise `statusLine` detection (no false-positive uninstall of user commands), `overrides` migration for legacy hook entries.\n- **1,152 tests** — +31 from v2.11.0 covering hooks adapter, tier manifest loader, override migration, and review regressions.\n\n### Previous (v2.11.0)\n\n- **Mesh integrity** — 8 dead wires fixed, 5 workflow gaps closed (hotfix chain, API versioning, monorepo mode, feature flags, dependency upgrade campaigns)\n- **audit v0.4.0** — DX Review Mode: Addy Osmani's 8 developer experience principles with scoring rubric and browser-pilot integration\n- **cook v2.4.0** — remediation cycle counter + upstream inconsistency protocol\n- **problem-solver v0.4.0** — Cynefin, SWOT, PESTLE, Porter's Five Forces, ethics framework\n- **plan v1.5.0** — autopilot suggested_next: autonomous execution path for Pro users after plan approval\n- **Autopilot routing** — skill-router Tier 1 entry for Pro autopilot (\"auto\", \"làm hết\", \"đi ngủ\" → autonomous mode)\n\n### Previous (v2.10.0)\n\n- **marketing v0.4.0** — anti-AI copy rules (banned phrases, 5 hook types, specificity mandate), expanded SEO audit with schema markup guide (10 types + `@graph` pattern), programmatic SEO awareness (4 playbooks), optional Pro content-scorer/cro-analyst integration\n- **Pro growth pack v1.1.0** — 3 new skills (content-scorer, cro-analyst, marketing-psych) + 6 existing skills enriched with SEO Machine patterns\n\n### Previous (v2.8.0)\n\n- **Anti-Loop Intelligence** — 7 core skills enriched with execution loop detection, saturation analysis, error pattern matching, artifact folding, budget-aware progression, and recovery policy routing\n- **cook v2.1.0** — observation/effect ratio tracking (detects stuck agents reading without writing) + budget-aware phase progression with hard caps on replans, quality retries, and session tool calls\n- **completion-gate v1.8.0** — execution loop audit: classifies tool calls as observation vs effect, flags imbalanced ratios and repeating sequences in gate reports\n- **scout v0.3.0** — info saturation detection: tracks entity discovery rate and content similarity to stop scanning when diminishing returns detected\n- **research v0.4.0** — diminishing returns detection: monitors new-entity ratio and result overlap across searches to skip redundant queries\n- **context-engine v0.9.0** — artifact folding: large tool outputs (\u003e4000 chars or \u003e120 lines) saved to `.rune/artifacts/` with compact preview in context\n- **debug v1.0.0** — known error pattern catalog: 8 error archetypes (STATELESS_LOSS, MODULE_NOT_FOUND, TYPE_MISMATCH, ASYNC_DEADLOCK, etc.) with recovery hints + error fingerprinting for dedup\n- **fix v0.8.0** — recovery policy matrix: classifies errors into 8 types (INPUT_REQUIRED→PROMPT_USER, TIMEOUT→RETRY, POLICY_BLOCKED→ABORT, etc.) before attempting fixes\n- **Source attribution cleanup** — removed all enrichment credit lines from skill files to reduce context noise\n\n### Previous (v2.7.0)\n\n- **Deep Knowledge** — 8 core skills enriched with battle-tested patterns: context compaction, structured cumulative memory, milestone analysis, multi-provider adapters, AI-driven interview, prompt-as-API-contract, token budget tracking, incremental stream processing\n- **946 Tests** — compiler + signals + hooks + scripts + status + visualizer validation\n\n### Previous (v2.6.0)\n\n- **Mesh Signals** — event-driven skill communication via frontmatter. Skills declare `emit` and `listen` signals; compiler builds a signal graph in `skill-index.json`. 23 signals across 15 core skills\n- **Signal Validation** — `scripts/validate-signals.js` checks orphan listeners (hard error), unlistened emitters (warning), signal naming conventions\n- **Mesh Contract** (v2.5.0) — `.rune/contract.md` project-level invariants enforced by cook + sentinel as hard gates\n- **Tier Override** — Pro/Business packs override Free packs with skill-level merging\n- **Scripts Bundling** — compiler copies `scripts/` directories, resolves `{scripts_dir}` placeholders\n\n### Signal Graph\n\nSkills communicate through declarative signals — no runtime event bus, just metadata for discovery, validation, and routing:\n\n```\nscout ──emit:codebase.scanned──→ plan, brainstorm\nfix ────emit:code.changed──────→ test, sentinel, review, preflight, verification\ntest ───emit:tests.passed──────→ deploy\ntest ───emit:tests.failed──────→ debug\nsentinel─emit:security.passed──→ deploy\ndebug ──emit:bug.diagnosed─────→ fix\ndeploy ─emit:deploy.complete───→ watchdog\ncook ───emit:phase.complete────→ session-bridge\n```\n\n## What Rune Is (and Isn't)\n\nRune started as a **Claude Code plugin** and now compiles to **every major AI IDE**. Same 63 skills, same mesh connections, same workflows — zero knowledge loss across platforms.\n\n| | Rune Provides | Claude Code Provides |\n|---|---|---|\n| **Workflows** | 8-phase TDD cycle (cook), parallel DAG execution (team), rescue pipelines | Basic tool calling |\n| **Quality Gates** | preflight + sentinel + review + completion-gate (parallel) | None built-in |\n| **Domain Knowledge** | 14 extension packs (trading, SaaS, mobile, etc.) | General-purpose |\n| **Cross-Session State** | .rune/ directory (decisions, conventions, progress) | Conversation only |\n| **Mesh Resilience** | 215+ skill connections, fail-loud-route-around | Linear execution |\n| **Cost Optimization** | Auto model selection (haiku/sonnet/opus per task) | Single model |\n| | | |\n| **Sandbox \u0026 Permissions** | — | Claude Code handles this |\n| **Agent Spawning** | — | Claude Code's Task/Agent system |\n| **MCP Integration** | — | Claude Code's MCP protocol |\n| **File System Access** | — | Claude Code's tool permissions |\n\n### Common Misconceptions\n\n| \"Rune doesn't have...\" | Reality |\n|---|---|\n| Task graph / DAG | `team` skill: DAG decomposition → parallel worktree agents → merge coordination |\n| CI quality gates | `verification` skill: lint + typecheck + tests + build (actual commands, not LLM review) |\n| Memory / state | `session-bridge` + `journal`: cross-session decisions, conventions, ADRs, module health |\n| Multi-model strategy | Every skill has assigned model: haiku (scan), sonnet (code), opus (architecture) |\n| Agent specialization | 62 specialized skills with dedicated roles (architect, coder, reviewer, scanner, researcher, BA, scaffolder) — each runs as a Task agent via Claude Code |\n| Security scanning | `sentinel`: OWASP patterns, secret scanning, dependency audit. `sast`: static analysis |\n\n## Install\n\n### Claude Code (Native Plugin)\n\n```bash\n# Install via Claude Code CLI\nclaude plugin add rune-kit/rune\n```\n\nOr add manually in `~/.claude/settings.json` under `installed_plugins`.\n\nFull mesh: subagents, hooks, adaptive routing, mesh analytics.\n\n### Cursor / Windsurf / Antigravity / Any IDE\n\n```bash\n# Compile Rune skills for your platform\nnpx @rune-kit/rune init\n\n# Or specify platform explicitly\nnpx @rune-kit/rune init --platform cursor\nnpx @rune-kit/rune init --platform windsurf\nnpx @rune-kit/rune init --platform antigravity\n```\n\nThis compiles all 63 skills into your IDE's rules format. Same knowledge, same workflows.\n\n### Platform Comparison\n\n| Feature | Claude Code | Cursor / Windsurf / Others |\n|---------|-------------|---------------------------|\n| Skills available | 61/61 | 61/61 |\n| Mesh connections | 215+ (programmatic) | 215+ (rule references) |\n| Workflows \u0026 HARD-GATEs | Full | Full |\n| Extension packs | 14 | 14 |\n| Subagent parallelism | Native | Sequential fallback |\n| Lifecycle hooks | 8 hooks (JS runtime) | Inline MUST/NEVER constraints |\n| Adaptive model routing | haiku/sonnet/opus | Single model |\n| Mesh analytics | Real-time metrics | Not available |\n\n**Same power, different delivery.** Claude Code gets execution efficiency; other IDEs get the same knowledge and workflows.\n\n## Quick Start\n\n```bash\n# Onboard any project (generates CLAUDE.md + .rune/ context)\n/rune onboard\n\n# Build a feature (full TDD cycle)\n/rune cook \"add user authentication with JWT\"\n\n# Debug an issue\n/rune debug \"login returns 401 for valid credentials\"\n\n# Security scan before commit\n/rune sentinel\n\n# Refactor legacy code safely\n/rune rescue\n\n# Full project health audit\n/rune audit\n\n# Respond to a production incident\n/rune incident \"login service returning 503 for 30% of users\"\n\n# Generate design system before building UI\n/rune design \"trading dashboard with real-time data\"\n\n# Bootstrap a new project from scratch (v2.1.0)\n/rune scaffold \"REST API with auth, payments, and Docker\"\n\n# Deep requirement analysis before building\n/rune ba \"integrate Telegram bot with trading signals\"\n\n# Auto-generate project documentation\n/rune docs init\n\n# Build an MCP server\n/rune mcp-builder \"weather API with forecast tools\"\n```\n\n## Auto-Discipline (Claude Code Hooks)\n\nTurn Rune skills into ambient runtime — no more `/rune preflight` every time. Install once, skills auto-fire on relevant tool calls:\n\n```bash\n# Wire Rune quality gates into Claude Code (.claude/settings.json)\nnpx @rune-kit/rune hooks install --preset gentle\n\n# Preset options:\n#   gentle  — advisory, never blocks (default)\n#   strict  — blocks tool call on BLOCK verdict\n#   off     — uninstall\n\n# Inspect current wiring\nnpx @rune-kit/rune hooks status\n\n# Remove (preserves user-authored hooks)\nnpx @rune-kit/rune hooks uninstall\n```\n\nWhat gets wired:\n\n| Event | Skill | When it fires |\n|---|---|---|\n| PreToolUse(Edit\\|Write) | preflight | Before Claude edits source files |\n| PreToolUse(Bash) | sentinel | Before shell commands (catches `git commit`, secrets) |\n| PostToolUse(Edit\\|Write) | dependency-doctor | After dependency manifest edits |\n| Stop | completion-gate | End of session — validates claims against evidence |\n\nRune only manages entries tagged with its command signature. User-authored hooks in the same events are preserved on install/uninstall.\n\n### Stacking paid tiers (Pro, Business)\n\nPaid tiers ship their own `hooks/manifest.json`. Point Rune at the install root and pass `--tier`:\n\n```bash\nexport RUNE_PRO_ROOT=~/rune-pro\nrune hooks install --preset gentle --tier pro\n\n# Stack Free + Pro + Business in one command\nexport RUNE_BUSINESS_ROOT=~/rune-business\nrune hooks install --preset gentle --tier pro --tier business\n```\n\nMulti-platform: tier hooks compile to Claude Code, Cursor, Windsurf, and Antigravity with the same command — no Claude-only lock-in.\n\n## Architecture\n\n### 5-Layer Model\n\n```\n╔══════════════════════════════════════════════════════╗\n║  L0: ROUTER (1)                                      ║\n║  Meta-enforcement — routes every action               ║\n║  skill-router                                         ║\n╠══════════════════════════════════════════════════════╣\n║  L1: ORCHESTRATORS (5)                                ║\n║  Full lifecycle workflows                             ║\n║  cook │ team │ launch │ rescue │ scaffold             ║\n╠══════════════════════════════════════════════════════╣\n║  L2: WORKFLOW HUBS (29)                               ║\n║  Cross-hub mesh — the key differentiator              ║\n║                                                        ║\n║  Creation:    plan │ scout │ brainstorm │ design │     ║\n║               skill-forge │ ba │ mcp-builder │ graft   ║\n║  Development: debug │ fix │ test │ review │ db         ║\n║  Quality:     sentinel │ preflight │ onboard │         ║\n║               audit │ perf │ review-intake │           ║\n║               logic-guardian                            ║\n║  Delivery:    deploy │ marketing │ incident │ docs     ║\n║  Rescue:      autopsy │ safeguard │ surgeon            ║\n║  Security:    adversary                                ║\n║  Velocity:    retro                                    ║\n╠══════════════════════════════════════════════════════╣\n║  L3: UTILITIES (27)                                   ║\n║  Stateless, pure capabilities                         ║\n║                                                        ║\n║  Knowledge:   research │ docs-seeker │ trend-scout     ║\n║  Reasoning:   problem-solver │ sequential-thinking     ║\n║  Validation:  verification │ hallucination-guard │     ║\n║               completion-gate │ constraint-check │     ║\n║               sast │ integrity-check                   ║\n║  State:       context-engine │ journal │               ║\n║               session-bridge                           ║\n║  Monitoring:  watchdog │ scope-guard                   ║\n║  Media:       browser-pilot │ asset-creator │          ║\n║               video-creator                            ║\n║  Deps:        dependency-doctor                        ║\n║  Workspace:   worktree                                 ║\n║  Git:         git                                      ║\n║  Documents:   doc-processor                            ║\n║  Security:    sentinel-env                             ║\n║  Memory:      neural-memory                            ║\n║  Packs:       context-pack                             ║\n║  Slides:      slides                                   ║\n╠══════════════════════════════════════════════════════╣\n║  L4: EXTENSION PACKS (14)                             ║\n║  Domain-specific, install what you need                ║\n║                                                        ║\n║  @rune/ui │ @rune/backend │ @rune/devops │            ║\n║  @rune/mobile │ @rune/security │ @rune/trading │      ║\n║  @rune/saas │ @rune/ecommerce │ @rune/ai-ml │        ║\n║  @rune/gamedev │ @rune/content │ @rune/analytics │    ║\n║  @rune/chrome-ext │ @rune/zalo                         ║\n╚══════════════════════════════════════════════════════╝\n```\n\n### Layer Rules\n\n| Layer | Can Call | Called By | State |\n|-------|---------|----------|-------|\n| L0 Router | L1-L3 (routing) | Every message | Stateless |\n| L1 Orchestrators | L2, L3 | L0, User | Stateful (workflow) |\n| L2 Workflow Hubs | L2 (cross-hub), L3 | L1, L2 | Stateful (task) |\n| L3 Utilities | Nothing (pure)* | L1, L2 | Stateless |\n| L4 Extensions | L3 | L2 (domain match) | Config-based |\n\n\\*L3→L3 exceptions: `context-engine`→`session-bridge`, `hallucination-guard`→`research`, `session-bridge`→`integrity-check`\n\n### Cost Intelligence\n\nEvery skill has an auto-selected model for optimal cost:\n\n| Task Type | Model | Cost |\n|-----------|-------|------|\n| Scan, search, validate | Haiku | Cheapest |\n| Write code, fix bugs, review | Sonnet | Default |\n| Architecture, security audit | Opus | Deep reasoning |\n\nTypical feature: ~$0.05-0.15 (vs ~$0.60 all-opus).\n\n## Key Workflows\n\n### `/rune cook` — Build a Feature\n\n```\nPhase 0 RESUME     → detect existing .rune/plan-*.md, load active phase\nPhase 1 UNDERSTAND → scout scans codebase, ba elicits requirements\nPhase 2 PLAN       → plan creates master plan + phase files\nPhase 3 TEST       → test writes failing tests (TDD red)\nPhase 4 IMPLEMENT  → fix writes code (TDD green)\nPhase 5 QUALITY    → preflight + sentinel + review (parallel)\nPhase 6 VERIFY     → verification + hallucination-guard\nPhase 7 COMMIT     → git creates semantic commit\nPhase 8 BRIDGE     → session-bridge saves state, announce next phase\n```\n\nMulti-session: Phase 0 detects existing plans and resumes from the current phase. One phase per session = small context = better code.\n\n### `/rune rescue` — Refactor Legacy Code\n\n```\nPhase 0 RECON      → autopsy assesses damage (health score)\nPhase 1 SAFETY NET → safeguard writes characterization tests\nPhase 2-N SURGERY  → surgeon refactors 1 module per session\nPhase N+1 CLEANUP  → remove @legacy markers\nPhase N+2 VERIFY   → health score comparison (before vs after)\n```\n\n### `/rune launch` — Deploy + Market\n\n```\nPhase 1 PRE-FLIGHT → full test suite\nPhase 2 DEPLOY     → push to platform\nPhase 3 VERIFY     → live site checks + monitoring\nPhase 4 MARKET     → landing copy, social, SEO\nPhase 5 ANNOUNCE   → publish content\n```\n\n## Mesh Resilience\n\nIf a skill fails, the mesh adapts:\n\n| If this fails... | Rune tries... |\n|---|---|\n| debug can't find cause | problem-solver (different reasoning) |\n| docs-seeker can't find docs | research (broader web search) |\n| scout can't find files | research + docs-seeker |\n| test can't run | deploy fix env, then test again |\n\nLoop prevention: max 2 visits per skill, max chain depth 8.\n\n## Cross-Session Persistence\n\nRune preserves context across sessions via `.rune/`:\n\n```\n.rune/\n├── decisions.md     — architectural decisions log\n├── conventions.md   — established patterns \u0026 style\n├── progress.md      — task progress tracker\n└── session-log.md   — brief session history\n```\n\nEvery new session loads `.rune/` automatically — zero context loss.\n\n## Extension Packs\n\nDomain-specific skills that plug into the core mesh:\n\n| Pack | Skills | For |\n|------|--------|-----|\n| @rune/ui | design-system, components, a11y, animation | Frontend |\n| @rune/backend | api, auth, database, middleware | Backend |\n| @rune/devops | docker, ci-cd, monitoring, server, ssl | DevOps |\n| @rune/mobile | react-native, flutter, app-store, native | Mobile |\n| @rune/security | owasp, pentest, secrets, compliance | Security |\n| @rune/trading | fintech, realtime, charts, indicators | Fintech |\n| @rune/saas | multi-tenant, billing, subscription, onboarding | SaaS |\n| @rune/ecommerce | shopify, payment, cart, inventory | E-commerce |\n| @rune/ai-ml | llm, rag, embeddings, fine-tuning | AI/ML |\n| @rune/gamedev | threejs, webgl, game-loops, physics | Games |\n| @rune/content | blog, cms, mdx, i18n, seo | Content |\n| @rune/analytics | tracking, a/b testing, funnels, dashboards | Growth |\n\n### Rune Pro — $49 lifetime\n\n\u003e *Free Rune makes Claude disciplined. Pro makes Claude self-aware.*\n\n**Context Intelligence** — the headline Pro feature. Claude Code auto-compacts at random, wiping your session context. With Pro, Claude **knows** when context is filling up — and proactively saves decisions, progress, and discoveries before compact hits.\n\n```\nFree:  Claude is blind to context pressure → auto-compact → amnesia\nPro:   Claude sees real-time context % → saves state → compact → reloads → zero loss\n```\n\nHow it works: `rune-pulse` (statusline) reads `context_window.used_percentage` from Claude Code, writes to temp file. `context-inject` hook injects warnings into Claude's context at 70% / 80% / 90%. Claude triggers `session-bridge` + `neural-memory` to persist everything. After `/compact`, `session-start` reloads `.rune/` state — Claude picks up exactly where it left off.\n\n**Autopilot** — approve a plan, walk away. Autonomous multi-session execution with zero-HIGH-tolerance quality gates, baseline regression checks, cross-phase coherence review, and structured completion reports. `cook` gets the job done. `autopilot` gets it done while you sleep.\n\n**Department Packs** — 34 skills across 5 domains:\n\n| Pack | What it does |\n|------|-------------|\n| **Product** | PRDs from user stories, roadmap prioritization, KPI dashboards, release comms, competitive analysis |\n| **Sales** | Account research briefs, call prep with objection handling, outreach sequences, pipeline health review |\n| **Data Science** | SQL exploration → visualization → statistical testing → ML eval, all in one flow |\n| **Support** | Ticket triage with SLA routing, KB article generation, escalation playbooks, support metrics |\n| **Growth** | Niche research, content scouting, SEO architecture, landing page generation, content health monitoring, data moats, content quality scoring, CRO psychology analysis, 74 marketing mental models |\n\nAll Pro packs plug into the core mesh — `cook` orchestrates them, `sentinel` gates them, `team` parallelizes them.\n\n**[Get Rune Pro](https://rune-kit.github.io/rune#pricing)** — [rune-kit/rune-pro](https://github.com/rune-kit/rune-pro)\n\n### Rune Business — $149 lifetime (includes Pro)\n\n\u003e *Pro handles departments. Business handles the company.*\n\nBusiness packs don't just add skills — they **wire departments together**. Finance pulls from sales pipeline. Legal audits product specs. Enterprise search indexes support KB. 40 cross-domain signals, zero manual context passing.\n\n| Pack | What it does |\n|------|-------------|\n| **Finance** | Budget planning from sales pipeline data, P\u0026L analysis, cash flow forecasting, compliance reporting |\n| **Legal** | Contract review with clause extraction, GDPR/SOC2 compliance checks, NDA triage, IP protection |\n| **HR** | JD generation, resume screening, structured interviews, comp benchmarking, onboarding workflows |\n| **Enterprise Search** | Cross-system knowledge retrieval with permission-aware filtering and knowledge graph |\n\n4 packs, 26 skills, 118 reference files, 11 automation scripts. Business includes all Pro features because it depends on Pro data — finance can't forecast without sales pipeline, legal can't audit without product specs.\n\n**[Get Rune Business](https://rune-kit.github.io/rune#pricing)** — [rune-kit/rune-business](https://github.com/rune-kit/rune-business)\n\n## Multi-Platform Compiler\n\nRune includes a 3-stage compiler that transforms SKILL.md files into platform-native rule formats:\n\n```\nskills/*.md → PARSE → TRANSFORM → EMIT → platform rules\n```\n\n**8 transforms applied per platform:**\n1. Frontmatter: strip Claude Code-specific directives\n2. Cross-references: `rune:cook` → `@rune-cook.mdc` (Cursor) / prose ref (Windsurf)\n3. Tool names: `Read`, `Edit`, `Bash` → generic language\n4. Subagents: parallel → sequential workflow\n5. Compliance: inject enforcement preamble (non-Claude platforms)\n6. Hooks: runtime hooks → inline MUST/NEVER constraints\n7. Branding: Rune attribution footer\n\n```bash\n# Build for any platform\nnpx @rune-kit/rune build --platform cursor\nnpx @rune-kit/rune build --platform windsurf\n\n# Validate compiled output\nnpx @rune-kit/rune doctor\n```\n\nSee [docs/MULTI-PLATFORM.md](docs/MULTI-PLATFORM.md) for the full architecture.\n\n## Documentation\n\n| Doc | What's inside |\n|-----|---------------|\n| [`docs/GETTING_STARTED.md`](docs/GETTING_STARTED.md) | Your first 5 minutes with Rune — install to first `/rune cook` |\n| [`docs/SKILLS.md`](docs/SKILLS.md) | All 63 skills, searchable by intent and layer |\n| [`docs/SIGNALS.md`](docs/SIGNALS.md) | Canonical signal inventory — 25 events, emit/listen graph |\n| [`docs/ARCHITECTURE.md`](docs/ARCHITECTURE.md) | 5-layer mesh architecture reference |\n| [`docs/VISION.md`](docs/VISION.md) | Philosophy — what Rune is and isn't |\n| [`docs/HOOKS.md`](docs/HOOKS.md) | Auto-discipline hooks per platform |\n| [`docs/TROUBLESHOOTING.md`](docs/TROUBLESHOOTING.md) | Common issues + fixes |\n| [`CONTRIBUTING.md`](CONTRIBUTING.md) | How to contribute skills, packs, fixes |\n| [`CHANGELOG.md`](CHANGELOG.md) | Release history |\n| [`ROADMAP.md`](ROADMAP.md) | What's next |\n\n## Numbers\n\n```\nCore Skills:       62 (L0: 1 │ L1: 5 │ L2: 29 │ L3: 27)\nExtension Packs:   14 free + 5 pro + 4 business\nMesh Connections:  215+ cross-references\nMesh Signals:      25 signals across 63 skills (emit/listen graph)\nConnections/Skill: 3.4 avg\nPlatforms:         8 (Claude Code, Cursor, Windsurf, Antigravity, Codex, OpenCode, OpenClaw, Generic)\nCompiler:          ~1400 LOC (parser + 8 transforms + 8 adapters + CLI)\nTests:             1,152+ (compiler + signals + status + visualizer + hooks + scripts + tier-hooks)\nPack Depth:        23 packs total (14 free + 5 pro + 4 business, all free packs rated Deep)\n```\n\n## Acknowledgments\n\n- **[UI/UX Pro Max](https://github.com/nextlevelbuilder/ui-ux-pro-max-skill)** (MIT, 42.8k★) — Design intelligence databases powering Rune's `design` skill and `@rune/ui` pack: 161 color palettes, 84 UI styles, 73 font pairings, 99 UX guidelines, 161 industry reasoning rules.\n\n## License\n\nMIT\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Frune-kit%2Frune","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Frune-kit%2Frune","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Frune-kit%2Frune/lists"}