broomva
UserBroomva agent-skills monorepo — 48 Tier-2 skills compatible with Claude Code, Codex, Cursor, Gemini CLI, Goose, Copilot. Layout follows anthropics/skills (agentskills.io spec). Install: npx skills add broomva/skills --skill <name>.
Categories
Indexed Skills (52)
livecoding
Algorave-grade livecoded music workflow — TidalCycles patterns (Haskell DSL driving SuperDirt over OSC) + Hydra-synth visuals (browser or VS Code Simple Browser via a local HTML page that loads hydra-synth from CDN). Wraps the boot ritual (SuperCollider → SuperDirt → Tidal → Hydra), a vibe-descriptor pattern generator (industrial, ambient, DnB, footwork, techno, algorave-glitch), and a reference tunnel-visuals page for the @lo.fi.sci.fi-projected aesthetic. Mac-first; ports to Linux with package-name swaps.
d1-cli
Shop Tiendas D1 (Colombia, d1.com.co) from the command line — search the catalogue, resolve your nearest physical store, price a basket against that store's real stock, and quote delivery. D1 runs VTEX IO (account `d1tiendas`), so this drives its public storefront API with no admin key at all — catalogue and cart work fully anonymously, and a one-time emailed code unlocks order history. Handles the two traps that make naive D1 automation wrong — availability is regionalized (an unregioned query reports a national catalogue nobody can actually buy from) and prices arrive in two different units (search reports whole pesos, checkout reports hundredths, a silent 100x). Builds and prices baskets; it deliberately cannot pay, handing a checkout URL to a human instead. USE WHEN the user wants to find D1 products or prices, check whether D1 delivers somewhere, build or cost a D1 grocery basket, compare D1 items, or review their D1 orders. NOT FOR other Colombian retailers (Éxito, Jumbo, Ara, Alkosto), and not for comp
procurer
Grounded procurement research for any real-world need. Turn a problem ("reduce bedroom noise from the avenue", "add air conditioning", "hire an accountant", "buy a TV mount", "replace the kitchen sink") into 3–5 ranked alternatives spanning DIY-retail → mid-retail → specialty → contractor → consultant/turnkey, each with cited price bands (low / typical / high), confidence scores, locale-aware currency, and a final recommendation with budget. Output is decision-shaped (recommendation + budget envelope), not research-shaped (knowledge artifact).
swapit
Stateful, local-first household toxics inventory + swap engine. Identify the items in a home that carry endocrine disruptors and persistent chemicals (BPA/BPS, phthalates, PFAS/PTFE, parabens, flame retardants, VOCs, microplastics), score each by *real* exposure (severity x presence x how it's used x condition), and track the swap to a safer alternative from "flagged" -> "sourced" -> "swapped". Ships a grounded, cited knowledge graph of ~20 hazards, ~40 item-classes, and ~40 alternatives. Hands sourcing off to the `procurer` skill. The skill's state is the source of truth — the agent is the app.
agentic-vps
Provision and harden a fresh Linux VPS into an autonomous-agent dev host using the capability-preserving model — the box IS the sandbox: full agent autonomy inside it (non-root sudo user, open egress, YOLO mode, any toolchain), contained by ISOLATION (non-root + VPN-only access + snapshot + no-long-lived-secrets), not by restricting the agent. Composes an idempotent provisioning script, a staged firewall/VPN lockout-safe sequence, and a deterministic verification gate. USE WHEN setting up a new VPS/cloud box for Claude Code / agentic development, hardening a machine to run coding agents unattended, "set up another machine like srv1692698", provisioning an agent dev host, or auditing whether a box meets the agentic-host security invariants. NOT FOR generic server hardening with no agent on it (that's standard CIS/Lynis), desktop/laptop setup, or Kubernetes/multi-tenant clusters. Triggers on "set up a vps for agents", "agent dev host", "harden a box for autonomous agents", "new agent machine", "provision agenti
brand-icons
Brand icon and visual identity management for BroomVA projects. Helps the agent produce favicons, app icons, Open Graph images, and social media avatars from brand assets, keeping a consistent visual identity across web, mobile, and social platforms. Use when: (1) generating favicon sets (ICO, PNG, SVG) for a web project, (2) creating app icons for iOS/Android, (3) generating Open Graph and Twitter Card images, (4) producing social media profile/banner images, (5) ensuring brand consistency across platforms.
design-engineering
Premium design engineering skill for agentic workflows — produces high-end, distinctive UI designs using DESIGN.md as the portable contract across Pencil MCP (in-IDE canvas), Figma MCP (team handoff + design tokens), and Google Stitch (vibe exploration + AI generation). Enforces anti-generic principles, WCAG 2.2 AA accessibility, perceptually uniform color, tiered animation timing, and screenshot-driven visual validation at every stage. Use when: (1) creating or reviewing UI designs from an agent, (2) generating DESIGN.md for a project, (3) implementing designs from Pencil .pen files or Figma URLs, (4) enforcing visual quality standards on AI-generated interfaces, (5) setting up a multi-tool design pipeline, (6) converting between design tools and production code. Triggers on: 'design engineering', 'design system', 'DESIGN.md', 'premium design', 'high-end UI', 'pencil design', 'figma to code', 'stitch design', 'vibe design', 'visual design review', 'glass design', 'design tokens', 'UI quality'.
tekton
Tekton — the shared architecture-intent substrate for co-designing systems with the agent. One typed graph across six tiers (system / journey / data / infra / decisions / qualities); views are queries, not separate diagrams. The canonical artifact is a diff-friendly YAML model both human and agent read and write; it renders to Mermaid (agent-legible, GitHub-native) and a self-contained tabbed HTML viewer (human-visual). Cross-tier traceability (`tekton query <from> <to>`) answers "which infra does this user-journey step touch?" as a path query. USE WHEN: designing or thinking deeply about architecture, a system, a data model, user journeys/flows, or a technical plan WITH the agent; when a Category-C HTML doc isn't enough because you need to see AND edit AND traverse the design across tiers; "let's design X", "architect this", "model the system", "draw the flow", "how does this fit together", "diagram this", "/tekton". NOT FOR: a one-off throwaway diagram (use Mermaid inline); prose-only ADRs (write the ADR);
dogfood
Per-bstack-P11 reflex 7+16 — explicitly trigger the Dogfood Plan + per-stack cookbook + Dogfood Receipt sequence. Auto-detects the tech stack from repo signals (Cargo.toml + src-tauri/ → tauri-sidecar; next.config.* → nextjs; app.json + expo → expo-rn; Cargo.toml solo → rust-cli; openapi.* or REST framework deps → rest-api; mcp.{json,yaml} → mcp-server), loads the matching cookbook entry from bstack/references/dogfood-patterns.md (broomva/bstack ≥ 0.13.0), produces a six-row Dogfood Plan (entry surface · driver · evidence · smoke · end-to-end · receipt anchor), optionally executes via the named surfaces (Interceptor mandatory for visual deploy verification; gstack, cliclick, screencapture, curl+jq, xcrun simctl compose per stack), and produces the final receipt with binary anti-rationalization check. Three subcommands: /dogfood (full sequence — plan → execute → receipt), /dogfood plan (plan only), /dogfood receipt (receipt only, assumes plan exists in PR body or AGENTS.md). Inherits from broomva/bstack: cookb
keel
Measure whether a codebase's verification actually touches the world, or whether it is checking itself. Keel gathers the verification edges it can read in a target (GitHub Actions, CircleCI, GitLab CI, Travis, package scripts, Makefile, Rakefile, pyproject and a dozen tool configs) — and reports the surfaces it recognised but could not parse, so blindness never passes for absence. It classifies each edge as anchored, self-referential, or unknown by asking whether the actor being verified can write to the signal's producer, and reports a grounding ratio. Novel cases are judged by the agent and then crystallized into probes — small reviewable scripts — so repeat shapes get cheaper every run. Then routes each ungrounded check to an anchored signal that already exists in the same graph. Use when: (1) auditing whether AI-generated or agent-maintained work is genuinely verified, (2) assessing how AI-native a codebase actually is — measured by what fraction of its verification the agents cannot author, (3) reviewing
health
Personal health knowledge graph — local-first ingest of Garmin (Apple Health, Whoop, Oura, CGM in v2+) traces into SQLite, projected to Obsidian daily-note frontmatter, synthesized into validated longevity-proxy metrics (HRV-CV, CTL/ATL/TSB, VO2max arc). Hex architecture so new sources drop in as adapters. NOT a coaching surface in v1.
bookkeeping
Universal knowledge engine — scores, promotes, and compounds knowledge across all sources into a permanent, query-able entity graph
kg
Load relevant entities from the bstack knowledge graph (research/entities/) for a given topic. Two-tier scoring: tier-1 (catalog-only, ~5ms — slug + tags + claim + links + sources match against docs/knowledge-index.md) is the fast first pass and the only stage for high-confidence exact-slug/tag queries; tier-2 (body-grep fallback, ~300ms — auto-fires when tier-1 returns < N matches OR its best hit is below the confidence floor, default 18) recovers topics whose vocabulary appears in entity prose but not the dense catalog (hit rate jumps from 70% → 100% on representative queries; this is hit rate ≥1, not true recall@N). Surfaces top-N entity bodies as a single context block the agent reasons over. This is a LOAD skill, not a query DSL — querying is what the agent does once loaded. Implements the LLM-as-index architecture (BRO-1223): substrate canonical, one projection (catalog) routes, agent IS the query engine. Empirical: peak per-query context drops from 29% → 4.6% of 1M (6.3× reduction); cumulative session
eve-forge
Forge a personalized eve agent for a business end-to-end — absorb the business's artifacts, author the eve agent/ dir, validate, deploy to Vercel (or a VPS), smoke-test against ground truth, register, and evolve. The deterministic core is three safety gates distilled from a driven benchmark (BRO-1677): deploy-safety (never ship auth:none), validate (eve info --json → 0 diagnostics + tools registered), and smoke (drive the deployed agent, assert vs a ground-truth example). USE WHEN building/onboarding an eve agent for a tenant, "forge an eve agent", "deploy an eve agent", "onboard <business> onto eve", or when the Claude-Code orchestrator/forge must turn absorption inputs into a running eve operator. NOT FOR benchmarking frameworks (that's a one-off), running the operator itself (the forge builds it; the operator runs cheap on eve), or non-eve agent frameworks.
what
Explain the CONCEPTS a session used, at the operator's register, grounded in what actually ran. After an agent works, you are handed results built out of vocabulary you did not choose — primitive short-names, entity slugs, library names, terms of art from the diff. `/what` names those terms, ranks them by what actually blocks understanding, defines each one, and anchors it to where it appeared and to its knowledge-graph entity. Fired mid-conversation with a short slice it degrades to the re-pitch case: "that did not land, say it again with more context and a simpler register." USE WHEN: /what, "wait what", "what did you just do", "explain that again", "I don't follow", "you lost me", "what do those terms mean", "re-pitch that", "explain the concepts", "what is all this jargon", after an autonomous arc or a long PR when the vocabulary outran the operator. NOT FOR: a summary of actions taken (that is the P1 Bridge session log); state for the next agent (that is /handoff); teaching a whole body of work to master
openrocket-sim
Headless rocket design, simulation, and optimization using OpenRocket's Java core engine. Includes rocket-sim CLI tool for batch simulations and parameter sweeps, plus an AI-driven optimization loop using Claude. Use when: (1) designing model rockets programmatically, (2) running headless flight simulations, (3) optimizing rocket designs for altitude/stability/velocity, (4) batch-processing .ork files, (5) parameter sweeps over launch conditions, (6) AI-driven design optimization, (7) building a simulation pipeline or API wrapper, (8) exporting designs for 3D printing or laser cutting, (9) user mentions 'openrocket', 'rocket simulation', 'flight sim', 'rocket design', 'model rocket', 'rocketry', 'apogee optimization', 'thrust curve', 'rocket-sim'.
sdr-satellite
Software-defined radio (SDR) and satellite reception toolkit — what to install, what you can hear from space, and how to compose the open-source stack (SatDump, SatNOGS, GNU Radio, rustradio, satkit). Covers hardware abstraction (SoapySDR, rtl-sdr-rs), DSP frameworks (GNU Radio, liquid-dsp, rustradio, radiorust), receiver UIs (SDR++, SDRangel, Gqrx, CubicSDR), satellite decoders (SatDump for NOAA APT, Meteor-M LRPT, GOES HRIT, Inmarsat), ground-station orchestration (SatNOGS network, sgoudelis/ground-station, Hamlib), orbit mechanics (SGP4 in Python + Rust, satkit astrodynamics, dSGP4 differentiable). Use when: (1) the user wants to receive satellite signals — weather imagery, ISS SSTV, telemetry, (2) building a ground station or scheduling observations on SatNOGS, (3) decoding TLEs and computing satellite passes, (4) composing SDR tooling into a Rust pipeline (e.g. for Life Agent OS, Opsis world-state events, or arcan/sensorium), (5) explaining what an RTL-SDR or HackRF actually does, (6) deciding between Sa
omnivoice
Local TTS, voice cloning, voice design, and video dubbing via the OmniVoice Studio MCP server (open-source ElevenLabs alternative; nothing leaves the machine, runs on MPS/CUDA/CPU). Use when: (1) generating speech from text in any of 646 languages, (2) cloning a voice from a 3-second reference clip, (3) designing a voice by gender/age/accent/pitch/style, (4) dubbing a video into another language, (5) listing voice profiles or personality presets, (6) producing narration where privacy, cost, or absent API keys matter, (7) non-English narration where Edge TTS/kokoro fall short, (8) batch audio for blog posts or content pipelines. Triggers: 'omnivoice', 'voice clone', 'clone this voice', 'tts', 'narrate', 'generate speech', 'voice synthesis', 'dub video', 'voice design', 'local tts', 'multilingual voice', 'narrate this post', 'elevenlabs alternative'.
drift-check
Compares stated priorities against where time and effort actually went, and produces a strategy drift report. Uses git log, vault priorities, and project docs to detect misalignment. Use when the user says "drift check", "am I on track", "priority alignment", "where did my time go", "strategy drift", or wants to audit focus vs intention.
stakeholder-update
Takes one set of project facts and generates three versions: technical for engineering, business-impact for leadership, and customer-facing for success teams. Use when the user says "stakeholder update", "write an update", "communicate this to different audiences", "translate for leadership", "customer-facing version", or needs to communicate the same information to different audiences.
weekly-review
Scans the vault for updates from the past week, surfaces what changed, and flags what needs attention. Produces a structured weekly review. Use when the user says "weekly review", "week in review", "what happened this week", "weekly summary", "end of week review", or at the end of a work week.
colab-remote
Orchestrate Google Colab Pro/Pro+ GPU instances as remote training backends via SSH. Compounds /agent-browser (to launch Colab sessions and install colab-ssh) with SSH (to operate the runtime remotely). Use when: (1) launching a Colab notebook for GPU training, (2) running training jobs on Colab from the local terminal, (3) transferring datasets/checkpoints to/from Colab, (4) monitoring GPU utilization on a Colab instance, (5) integrating Colab GPU compute with /autoany EGRI optimization loops, (6) reconnecting to a Colab session after timeout. Triggers on: "colab", "colab-remote", "colab ssh", "colab training", "remote GPU", "colab pro", "train on colab", "google colab".
remote-gpu
Orchestrate a headless GPU server (NUC, cloud VM, or any SSH-accessible machine) from a local Mac or workstation. Use when: (1) Running GPU workloads remotely (training, inference, video generation), (2) Triggering Claude Code sessions on a remote machine, (3) Launching autoany EGRI loops, symphony orchestrations, or training scripts on remote hardware, (4) Managing jobs on a headless GPU server (submit, monitor, cancel, download results), (5) Setting up SSH tunnels or API bridges to a GPU machine, (6) Orchestrating multi-machine agent workflows. Triggers on: remote gpu, headless server, gpu server, remote training, remote inference, ssh gpu, nuc server, remote claude code, remote agent, gpu orchestration.
arcan-glass
BroomVA trademark web styling system — Arcan Glass design language for Next.js + Tailwind v4 + shadcn/ui projects. Use when creating or styling BroomVA web interfaces, implementing glass/frosted UI effects, setting up brand-consistent dark-first themes, or integrating ArcanBG brand tokens (AI Blue
design-distill
Distill a visual style from reference sites/products the user likes, then articulate it into a validated, dual-mode design system and a ready-to-run Claude Design ("Create Design System") handoff. Crystallized from the agents-dashboard run (checkit on handrovermeulen/Obsidian-Dashboard → moodboard → dual-mode tokens → Claude Design brief). It is a COMPOSITION skill — it fires existing tools (agent-browser capture, the Obsidian-Dashboard moodboard method, P18 format discipline, P11 dogfood validation, arcan-glass / ui-ux-pro-max) in sequence; it does not reimplement them. Core principle: MEASURED, NOT INFERRED — the palette and type come from computed styles extracted via headless browser, never from the model guessing at a screenshot. USE WHEN: distill a style, capture references, build a moodboard, "I like the look of X / use these as references", extract a palette from a site, design tokens from screenshots, create a design system from references, dual-mode (dark+light) design system, make a Claude Design b
finance-substrate
Personal finance and tax management substrate for Colombian residents. Imports bank transactions, tax certificates, and DIAN data. Projects annual tax liability (Form 210) with certificate-level accuracy. Handles password-protected PDFs from Colombian banks. Zero external paid services — all data stays local.
haima
Agentic finance engine for the Agent OS — x402 machine-to-machine payments, on-chain settlement, per-task revenue billing, and wallet management. Integrates with Arcan (agent runtime), Autonomic (economic homeostasis), and Lago (event-sourced persistence) to give agents real financial agency. Use when: (1) adding payment capabilities to an agent or agent-controlled API, (2) implementing x402 protocol support for HTTP-native machine payments, (3) wiring per-task billing so agents can charge for completed work, (4) managing agent wallets (secp256k1 keypair, encrypted storage), (5) configuring payment policies (auto-approve, spend caps, rate limits), (6) integrating Coinbase CDP or self-hosted facilitators for on-chain settlement, (7) bridging on-chain USDC with Autonomic's micro-credit state. Triggers on 'agent payments', 'x402', 'machine payments', 'agentic finance', 'haima', 'payment port', 'wallet management', 'per-task billing', 'USDC payments', 'agent revenue', '402 payment required'.
agentic-control-kernel
Unifying control-systems metalayer for LLM-as-controller agent development. Bootstrap any repository with typed plant/action/trace schemas, safety shield conventions, multi-rate loop hierarchy, EGRI-compatible evaluator interfaces, and the full consciousness stack (governance + knowledge graph + episodic memory). Use when: (1) setting up agentic control primitives in a new or existing project, (2) designing LLM-as-controller architectures with safety shields and typed directives, (3) wiring EGRI/autoany loops for controller or artifact improvement, (4) bootstrapping the consciousness stack (control-metalayer + knowledge-graph + conversation bridge), (5) integrating symphony-style orchestration patterns, (6) defining plant interfaces, state estimators, or world models for agent-controlled systems, (7) user says "control kernel", "agentic control", "safety shield", "plant interface", "control metalayer", "agent controller", "multi-rate loop", "LLM control law".
architecture-design-principles
Distilled architecture & design principles for building self-service developer platforms, control-plane / data-plane separation, and edge-centralized cross-cutting concerns — drawn from a senior platform engineer's 8-year retrospective on building Atlassian's Envoy-based load-balancing platform (Open Service Broker + Sovereign xDS control plane + AWS infra + sidecar model). Covers the three IaC layers (image / infra / runtime), the template+context pattern for dynamic proxy configuration, forced-migration as an adoption tool, long-term maintenance discipline (churn as a smell, build vs maintain as distinct skills), and the non-technical engineering skills (diplomacy, mentoring vs teaching). Use when: (1) designing a self-service internal developer platform, (2) deciding between native proxy features vs sidecars, (3) introducing or migrating to Envoy / xDS / service mesh, (4) implementing the Open Service Broker spec or any resource-provisioning API, (5) abstracting cloud complexity (AWS CloudFormation, Packer
bstack
bstack primer — the Broomva Stack: twenty irreducible automation primitives (P1–P20) that turn an agent-driven workspace into a self-operating system. This is the AGENT-READABLE PRIMER (primitive contract + short-name index + when each fires), NOT the CLI. The bstack CLI + governance substrate (bin/, doctor, templates, hooks) is a separate product installed by clone + bootstrap (git clone https://github.com/broomva/bstack && ./bin/bstack bootstrap) — do NOT `npx skills add broomva/bstack` (it drops bin/scripts, the repo-root SKILL.md case of vercel-labs/skills#1523). Use this primer to reason with the primitives: Bridge (P1), Gate (P2), Tickets (P3), Pipeline (P4), Fanout (P5), Bookkeeping (P6), Freshness (P7), Janitor (P8), Wait (P9), Hygiene (P10), Empirical (P11), Persist (P12), Dream (P13), Dep-Chain (P14), Snapshot (P15), Crystallize (P16), Lens (P17), Audience (P18), Orchestrate (P19), Cross-Review (P20). Triggers on 'bstack', 'Broomva Stack', 'the primitives', 'P1'..'P20', 'primitive contract', 'bstack
cross-review
bstack P20 — Cross-Model Adversarial Review Gate. The model that wrote the code cannot be the final judge of the code. Before substantive PRs merge, fire a cross-model adversarial gate — different evaluator than writer, anti-slop scoring ≥7/10, max 3 fix rounds, verdict logged in PR. Three strata: (A) Codex CLI cross-vendor for true different-model verdict, (B) fresh-context subagent under devils-advocate brief, (C) composed existing adversarial-review skills always parallel. Use cross-review when: (1) about to push a substantive PR (>200 LOC OR public API OR multi-file OR governance-class), (2) reviewing a draft plan/design before implementation, (3) auditing a feature spec against single-model blind spots, (4) integrating with the /autonomous skill's pre-push gate. Triggers on 'cross-review', 'P20', 'adversarial review', 'anti-slop', 'cross-model gate', 'different evaluator', 'devils advocate gate', 'self-review prohibition'.
harness-engineering-playbook
Implement OpenAI Harness Engineering practices in any repository. Use when setting up or refactoring agent-first workflows, writing or upgrading AGENTS.md and PLANS.md, creating deterministic smoke/test/lint/typecheck harness commands, defining strict architecture boundaries and data-shape contracts, wiring observability from day 1, and adding entropy-control checks plus CI automation for reliable autonomous runs.
unhobble
Audit and rightsize a context surface — CLAUDE.md, AGENTS.md, a SKILL.md, a system prompt, or a single prompt — against the Claude-5 context-engineering reversals. Measures the surface (token budget, how much of it is phrased as hard rules vs judgment, near-duplicate sections, contradiction candidates, filesystem-derivable content, progressive-disclosure gaps), then adjudicates each section keep / relocate / delete using a grounding-aware predicate: prose backed by an independent mechanism is decorative and free to cut, prose that is the only carrier of a behavior is a real bet. Also rewrites prompts and skills from prohibition-form into judgment-form. Anthropic deleted >80% of Claude Code's system prompt for Claude 5 with no eval loss; this is the repeatable version of that cut. Use when: (1) a CLAUDE.md or AGENTS.md has grown and you want to know what is safe to remove, (2) an agent is ignoring or colliding with its own instructions, (3) a prompt is long and over-constrained and underperforming, (4) auditin
founder-mode-oncology
Personalized cancer treatment navigation — maximal diagnostics, parallel therapy, therapeutic development, structure-based protein design. Use when (1) researching cancer treatments or combinations, (2) designing diagnostic strategies, (3) evaluating neoantigen vaccines or radioligand therapies, (4) navigating FDA expanded access or IRB processes, (5) interpreting ctDNA/MRD data, (6) building bioinformatics pipelines, (7) scRNA-seq target discovery, (8) AlphaFold/RFdiffusion/ProteinMPNN for neoantigen validation or de novo binder design, (9) patient advocacy. Keywords — cancer, oncology, neoantigen, vaccine, immunotherapy, radioligand, ctDNA, scRNA-seq, FAP, AlphaFold, RFdiffusion, ProteinMPNN, personalized medicine.
braindump
Takes raw unstructured thoughts, voice transcript dumps, or stream-of-consciousness text and auto-files them into the right Obsidian vault folders with tags, backlinks, and proper frontmatter. Use when the user says "braindump", "brain dump", "capture this", "dump my thoughts", "I'm thinking about...", or provides a long unstructured block of text to organize.
comprehend
Agent→human teach-to-mastery loop. Turns the agent into a wise, effective teacher whose single objective is that YOU deeply understand a body of work — the problem and why it existed, the solution and why it was resolved that way (design decisions, edge cases), and the broader context (why it matters, what it impacts). Stage-gated (confirm mastery before advancing), active-recall driven (you restate first; the agent fills gaps; quizzes via AskUserQuestion), and goal-bounded (the session does not end until your understanding is verified). Default subject = the current session/diff; pass a PR number, file, or subsystem to teach that instead. The pedagogical INVERSE of grill-me: grill-me stress-tests YOUR forward plan; comprehend transfers mastery of EXISTING work to you. Distinct from handoff (agent→agent narrative) and Bridge/Bookkeeping (agent→KG). Use when: (1) you want to deeply understand what was just built, (2) onboarding yourself or a teammate onto a PR / file / subsystem, (3) post-mortem learning after
claude-code-channels
Set up Claude Code messaging channels for Telegram and Discord — bot creation, plugin installation, token configuration, access control (pairing, allowlists, guild channels), and troubleshooting. Use when the user asks to connect Claude Code to Telegram or Discord, configure a messaging bot, set up channel access, enable guild channels or threads, troubleshoot MCP server failures, or manage channel permissions. Triggers on "telegram channel", "discord channel", "claude code channels", "messaging bot", "set up telegram", "set up discord", "pairing code", "discord bot", "telegram bot", "channel access", "--channels".
claude-remote-sessions
Per-channel remote sessions for Claude Code via Discord and Telegram — each channel, thread, or chat gets its own isolated Claude Code session via tmux, with per-channel access control, project-specific workdirs, and automatic CLAUDE.md chain loading. Includes session managers, watchdog daemons (auto-respawn, auto-discover, stale cleanup), thread context injection, session resume, workdir mapping, and launchd boot persistence. Use when: (1) setting up per-channel Discord/Telegram sessions for Claude Code, (2) managing multiple sessions across messaging channels, (3) auto-discovering new channels or threads, (4) spawning thread sessions with parent conversation context, (5) keeping remote agent sessions alive, (6) cleaning up stale sessions, (7) setting up Telegram per-chat sessions. Triggers on 'remote sessions', 'discord sessions', 'telegram sessions', 'per-channel discord', 'discord watchdog', 'telegram watchdog', 'spawn discord session', 'claude remote', 'channel session', 'thread session'.
bitnet
Microsoft BitNet — 1-bit LLM setup, inference, and benchmarking on CPU. Automates the full workflow: clone bitnet.cpp, create conda env, download GGUF models from HuggingFace, build optimized ternary kernels, and run inference. Supports official Microsoft models (2B) and community models (0.7B-10B). Use when: (1) setting up BitNet/bitnet.cpp for local CPU inference, (2) downloading and running 1-bit/ternary LLMs, (3) benchmarking BitNet vs full-precision models, (4) building edge/agentic inference pipelines without GPU, (5) converting HuggingFace models to GGUF for bitnet.cpp. Triggers on: 'bitnet', '1-bit llm', '1.58-bit', 'ternary model', 'ternary weights', 'edge inference', 'cpu inference', 'bitnet.cpp', 'bitlinear', 'no gpu inference'.
heretic-abliteration
Heretic — fully automatic LLM censorship removal (abliteration) and serving the result via Ollama. Wraps p-e-w/heretic: directional ablation (difference-of-means refusal directions, orthogonalizing attn.o_proj + mlp.down_proj) with an Optuna/TPE search that co-minimizes refusal count and KL divergence. Automates the full local workflow with the dependency + device fixes that make it actually run: clean-venv install (+ kernels/jinja2 fixes), device-correct run (CPU on Apple Silicon — MPS is blocked by a PyTorch GQA matmul bug; CUDA for real models), and the HF→GGUF→Ollama bridge. Includes a no-compute fast path: pull a pre-abliterated GGUF straight into Ollama. Use when: (1) decensoring / uncensoring an open-weight LLM locally, (2) running heretic-llm and hitting install or MPS/CUDA errors, (3) converting a Heretic/abliterated HF model to GGUF and serving it via Ollama CLI, (4) pulling a ready-made abliterated model into Ollama, (5) understanding abliteration as an evaluator-governed (EGRI) optimization loop.
tribe-v2-agent-alignment
Use Meta's TRIBE v2 brain encoder to validate cortical alignment of AI model representations (LLaMA, V-JEPA2, Wav2Vec, or any encoder) and inform model selection in the Life/Arcan agent OS stack. Use when: (1) Benchmarking whether a new model encoder aligns with human cortical processing, (2) Comparing text encoders by language cortex alignment score, (3) Comparing video encoders by visual cortex alignment score, (4) Integrating neuro-alignment scores into Arcan model routing, (5) Validating that a fine-tuned model has not lost biological plausibility, (6) Selecting the most brain-aligned encoder for a given modality in the agent OS, (7) Any task involving neuroscience-informed AI model evaluation, cortical alignment benchmarking, or biologically-inspired model selection.
tribe-v2-bci-applied
Applied BCI research and neuro-informed content optimization using Meta's TRIBE v2 brain encoder. Predicts neural responses to media, UI, and content without brain scanners — enabling stimulus optimization, attention ranking, and BCI groundwork research. Use when: (1) Predicting neural responses to video, audio, or text content, (2) Ranking content by predicted brain engagement (visual cortex, attention, emotion), (3) Optimizing stimuli to maximize activation in a target cortical region, (4) BCI research using non-invasive population priors, (5) Accessibility research (testing presentation formats for language/auditory processing), (6) Computational neuromarketing research (CC BY-NC, non-commercial only), (7) Any task involving brain response prediction for media, UI design, or BCI applications using TRIBE v2.
tribe-v2-neuroscience
In-silico neuroscience experiments using Meta's TRIBE v2 (TRansformer for In-silico Brain Experiments). Predicts fMRI cortical responses to video, audio, and text without brain scanners. Use when: (1) Designing and running virtual neuroscience experiments, (2) Predicting brain responses to stimuli, (3) Replicating classic neuroscience paradigms computationally, (4) Mapping which stimuli activate specific cortical regions (visual, auditory, language, DMN), (5) Testing hypotheses about neural processing before expensive fMRI studies, (6) Generating synthetic fMRI data for research, (7) Any task involving brain response prediction, cortical mapping, or computational neuroscience using TRIBE v2.
autonomous
Use when the user has agreed on a plan or selected from suggested options and wants the agent to execute the work autonomously without further instruction. Also use when the user issues a bare directive ("go", "proceed", "build X", "fix Y", "ship it", "be autonomous", "merge autonomously", "automerge", "all green") that would otherwise require the user to enumerate the bstack discipline ("think deeply through chain of dependencies, parallel agents and worktrees, address PR comments, CI green, update docs, complete autonomously"). Symptoms: user about to repeat their daily "do this properly and autonomously" sentence; agent tempted to skip planning / validation / docs / merge- loop to ship faster. Replaces a ~50-word ritual instruction with one invocation. Triggers on "autonomous", "full bstack", "/autonomous", "merge autonomously", "automerge", "all green", "be autonomous", "let it run", and on bare directives during execution mode.
governed-autonomy-loop
Turn any work-queue + any enforcement pipeline into a self-driving, self-healing, human-minimal autonomy loop with a control-systems safety envelope. A prose-defined governor closes a feedback loop over the queue, drives isolated worktree arcs to Done, is metacognitive (defers when the runtime can't support the act, escalates only when genuinely blocked, self-heals when the block clears), and NEVER performs the irreversible act itself — it delegates every merge/deploy/publish to an isolated arc gated by adversarial review + a policy gate + a server-side aggregate check. The controller is a markdown prompt interpreted fresh each tick; a thin tested shell (tick.sh) only schedules, locks, and logs. The skill SCAFFOLDS + GOVERNS a new loop instance: the invariant spine stays fixed, four adapters (tracker / irreversible-act + enforcement / runtime / partition) are per-instance, the deterministic core (scheduler, in-flight fold, reseed gate, config + denylist validation) is extracted into tested scripts, and the la
unslop
Remove the "vibecoded" / AI-slop look from an arbitrary frontend codebase at the ROOT, autonomously, and prove the result clears a crafted floor. Full-repo traversal (any framework) → root-cause attribution (the one file that declares each default: font, icons, color/radius/shadow tokens, gradient/glass, motion, copy voice) → fix once at the root, never at forty call sites → the substance tells no detector covers (legal routes, loading/empty/error states, placeholder and fake content, real product evidence) → render at 1280+390 → deterministic gate (scripts/unslop_gate.py, exit 0 = clear). Composes the impeccable detector (64 rules) and commands, Anthropic frontend-design for direction, broomva-design as the house example; reimplements none of them. USE WHEN: a site or app "looks vibecoded", "looks AI-generated", "looks like every other AI site", "feels generic/templated/slop", "make it feel premium / crafted / intentional", "de-slop / unslop this", "does this pass the vibecoded test", or before shipping any
goodies
Ingest, contextualize, and index curated resources (websites, tools, design references, research papers, UI kits) into a public GitHub Pages vault while automatically linking them to the local Knowledge Graph via /checkit and P6 bookkeeping. Dynamically resolves and expands category ontologies, generates client-side search indices, and publishes the updated vault database to a public repository. USE WHEN: /goodies, goodies, goodies add, add to goodies, bookmark this site, save to goodies, bookmark good resource, curate website. NOT FOR: internal-only notes without public vault indexing; general web searches without curating a resource.
morning-briefing
Reads open action items, this week's priorities, and recent vault updates, then produces a focused "start your day" brief. Use when the user says "morning briefing", "daily brief", "what's on my plate", "start my day", "what should I focus on today", or at the start of a work session.
skills-showcase
Generate a polished Remotion video and X thread showcasing the full agent skills inventory. Use when creating social media content about skills, rendering category-based skill visualizations, or producing animated showcases of agent capabilities. Triggers on "skills video", "showcase skills", "skills thread", "render skills", "social content for skills", or requests to visualize the skills inventory.
ccr
ccr — reversible payload compression: shrink any blob (tool output, log, RAG chunk, file) BEFORE it enters an LLM's context, while caching the full original locally so it can be expanded byte-for-byte on demand. The payload-axis counterpart to a knowledge-graph loader (the retrieval axis): the model sees a compact lossy view + a handle (ccr://<sha256>), and calls retrieve(handle) only when it needs the bytes back. Content-routed deterministic compactors (json skeleton / code outline / text head-tail); stdlib-only, no ML. Lifted from the CCR component of github.com/chopratejas/headroom. USE WHEN — a tool output / log / RAG chunk / file is too large for context and you want to compress it reversibly; "compress this payload", "shrink this before the model", "reversible compression", "cache the original and give me a handle". NOT FOR — loading knowledge-graph entities (that's the kg loader, the retrieval axis); lossless whole-file compression (use gzip); semantic summarization that does not need byte-exact recove
colombia-conflict
Knowledge engine over the Colombian Truth Commission (CEV) final report "Hay Futuro Si Hay Verdad" (2022) — the findings, statistics, armed-actor responsibilities, differential harms, coined lexicon, and the 67-recommendation roadmap for non-repetition (no repetición). Uses a kg/LLM-wiki retrieval architecture (substrate-canonical knowledge pages + a dense catalog projection for tier-1 routing + body-grep tier-2 + the agent as query engine) plus data-backed tools, including an `align` check that scores a proposed policy/action against the report's recommendation roadmap. USE WHEN reasoning about the Colombian armed conflict, peace process, transitional justice, victims, reconciliation, paramilitarism, the FARC-EP/ELN, narcotrafficking and drug policy, land/agrarian questions, the enemigo interno, falsos positivos, the Unión Patriótica genocide, social-leader killings, differential harms (ethnic peoples, women and LGBTIQ+, children, exiles, campesinado), territorial/human security, memory, or non-repetition; w
broomva-design
Materialize and apply Broomva's platform-neutral calm blue-axis design system across digital products. Includes a portable DESIGN.md contract, semantic styling foundations, general React/web primitives, product-pattern and platform-adaptation guidance, and an optional agentic-work extension. Use for Broomva-branded web, native mobile, desktop, commerce, content, analytics, SaaS, internal tools, communication products, design reviews, or when a request mentions Broomva design, the blackhole mark, blue-axis styling, Undertow, receipts, DESIGN.md, or the Broomva component library.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.