workflow-kit

The front door to the Pipeline Coordination system. One guided skill that doubles as installer + onboarder + meta-plan builder + first-handoff drafter + launcher. It scans a project's recent work for a goal and a candidate work-pile, sorts that pile WITH the operator, writes a coordination meta-plan PL and a first delegation handoff (HAN), installs the handoff-handling overlays into the create-note and pickup skills, teaches the three roles (orchestrator / coordinator / worker) in plain words, and prints exact launch instructions. Trigger on "meta-plan-init", "/meta-plan-init", "set up coordination", "set up the meta-plan", "bootstrap the pipeline coordination", "onboard me to the coordination system", "adopt my existing work into a meta-plan", or any request to stand up the coordinated-pipeline-execution layer for a project. Messy-work-adopt is the primary path; clean-start is the degenerate case.

Coordinator playbook for executing a delegation handoff issued by a meta-plan orchestrator. Auto-invoked by /pickup when it detects a delegation HAN (a `HAN - *.md` with `category: Handoff`, an `orchestrator_contact`, a `related_pl` pointing at a meta-plan, and `## Checkpoint log` + `## Clarification requests` sections) and this session is executing it. Also trigger when the operator says "coordinate this handoff", "I'm the coordinator", "run this HAN", "work the delegation". The coordinator is the top-level interactive instance the operator chats with: it stands up a workgroup of meta-plan-worker subagents, validates their deliverables, and posts LEAN progress to the HAN for the orchestrator. It does not own or edit the meta-plan.

Get oriented in the current project or Obsidian vault — read agent configs, guardrails, vault structure, and lessons learned. Use this skill at the start of any new session, after context compaction, when entering an unfamiliar project directory, or when the user says "orient", "get oriented", "what's the context here", "read the agents file", or "start up". Also trigger on /orient with no arguments. Even continued sessions count as new starts after compaction — always re-read the config chain.

Run a structured open-source landscape survey before designing or rebuilding a complex system. Six phases: bracket, discover (video + oracle + grounding doc), workspace (scratch clone), anchor read (manual, 1-2 closest peers), parallel deep-reads (team-dispatched), synthesis (single-worker comparative RE). Produces per-repo AREs + one comparative-analysis RE + grounding-doc cross-ref appendix; hands off to SD update and SPC drafting. Use whenever the user says "landscape survey", "deep-read these repos", "what should we borrow from X", "research before we spec", "evaluate candidates before the rebuild", "comparative source analysis", "what's the state of the art for [domain]", or invokes /landscape-survey. Also fires when the user points the assistant at a candidate set (a list of repos / libraries / approaches) and asks for a build plan before the SPC. The skill is the structural answer to "we built the system on marketing-page understanding and only discovered the failure mode after shipping."

Log work to the daily note AND project log. Use this skill when the user wants to record what they've been working on, says "log this", "log work", "add to worked on", "update the daily note", or wants to capture implementation progress. Always writes to both layers: daily note (human-readable) and PJL (agent-readable). Also trigger on "what did I work on" if the user wants to retroactively log work.

Load pickup documents (PICs) and resume work. Handles both single-PIC loading and full backlog triage. Use this skill when the user says "pickup", "pick up", "start of day", "what's on my plate", "what do I need to do", "get started", "load the pickup", "continue where I left off", "resume work", "triage pickups", "which pickups are easiest", "work through my pickups", "clear out pickups", "pickup backlog", "how many open pickups", "rank my pickups", "what's the quickest pickup to close", "batch my pickups", "group pickups by project", or points at a specific PIC document. Also trigger when the user opens a new session and says something like "ok what am I doing today" or "what was I working on".

Execute implementation plans with minimal ceremony — dispatch workers, track progress inline in the plan file. No external PM tools. Redesigned based on Anthropic's agent team research: default to the simplest pattern that works, replace agent-based QA with automated checks, eliminate PM relay. Use this skill when the user has a ready plan (PL - *.md) and wants to start building, resume a partial implementation, or when create-plan has just finished. Trigger on "implement", "build this", "start the plan", "kick off", "resume implementation", "let's go", "execute", or the user pointing at a plan file. Third step in the spec>plan>implement pipeline.

Run a retrospective — audit what happened, analyze what worked and what didn't, and apply process improvements. Works for sprints, phases, milestones, or any bounded unit of work. Also generates the handoff section (carry-forwards, infra changes, Plane refs) so the next session can resume efficiently. Use after completing a sprint or phase, when the user says "retro", "retrospective", "sprint review", "what went well", "what went wrong", "what can we improve", "review the sprint", or wants to analyze the build:fix ratio and prevent recurring issues. Also trigger automatically at the end of any /implement sprint that had incidents (fix cycles, user-caught bugs, compliance failures). Even casual requests like "let's look at what happened this sprint" or "how did that go" should trigger this skill when an implementation sprint just completed.

Sync core Workflow Kit skills and conventions to/from the public GitHub repo. The distribution channel for the WFK. Use this skill when the user says "update wfk", "update kit", "sync kit", "push skills", "pull skills", "update-wfk", "update-skills", "update-workflow-kit", or wants to push improvements to the WFK repo or pull the latest kit updates. This skill replaces the old /update-skills and /update-workflow-kit commands.

Memory and lesson system consolidation — scan automemory files, merge duplicates, resolve contradictions, prune stale data, cross-reference feedback memories against project lessons for supersession, sync any lesson trigger index, flag stale project lessons, and tighten the MEMORY.md index. Use this skill whenever the user says 'dream', 'consolidate memory', 'clean up memory', 'prune memory', 'memory consolidation', '/dream', or mentions that their memory feels bloated, stale, or contradictory. Also trigger when the user is doing end-of-day closeout and wants a memory hygiene pass, or says 'tighten memory', 'memory cleanup', 'dedupe memory'. Even a casual 'my memory is getting messy' should trigger this.

Extract and store reusable lessons or enforceable rules from the current conversation. Use this skill whenever the user says "learn this", "remember this", "don't do that again", or after any correction, failure, surprising fix, or non-obvious workaround worth capturing for future sessions. Also trigger when the user says "learn", "add a lesson", "add a rule", "that should be a rule", or describes a mistake they don't want repeated. Even if the user just says "/learn" with no arguments, this skill should activate and scan the conversation for the most recent teachable moment.

Three-role pipeline coordination: Worker implements, Verifier tests with I/O proof, Coordinator orchestrates. Prevents premature completion by ensuring no agent both implements and evaluates its own work. Use this skill for complex pipeline work where data quality matters (signal pipelines, entity creation, corpus processing, any LLM-in-the-loop system). Trigger on "qa-coord", "coordinate", "worker-verifier", or when starting pipeline evaluation/implementation work that has previously suffered from premature completion claims.

Pre-implementation spec check. Auto-invoke before building any feature on an existing system. Scans the project's specs directory, flags unimplemented specs in the same domain, blocks work until specs are acknowledged. Prevents reimplementing what's already been designed.

Open any non-trivial work session by writing a printed contract that bounds the work: surface, success criteria, anti-scope, validation plan, handoff trigger. The contract is referenced at every stage gate so the agent cannot drift into scope creep without consciously deciding to. Use this skill at the start of any non-trivial work, when the user says "bracket this", "set the bracket", "scope this", "what's the bracket", before /implement, before any multi-step code change, when picking up a critical PIC, or whenever a session is about to do real work that could expand. Also trigger when the user says /bracket. Even casual phrasing like "scope this work" or "what are we actually doing here" should trigger this. The skill is the structural answer to the failure mode "agent read the working contract and then violated it within hours."

End-of-session closeout — log work to the daily note and create pickup docs (PIC) for the next workday. Use this skill when the user says "closeout", "close out", "wrap up", "end of day", "EOD", "wind down", "done for the day", "call it a day", "closing out this session", or wants to finalize a work session and set up tomorrow's context. Also trigger when the user says "create a pickup", "make a PIC", or wants to carry forward work context to the next session. Even casual requests like "ok I'm done" or "wrapping up" at the end of a work session should trigger this skill.

Unified vault document creator. Handles SD, SPC, PIC, MN, PD, FD, DD, SO, RE, EB, and CAM documents with correct frontmatter, routing, writing profiles, and type-specific structure. Replaces individual create-sd, create-spec, create-pickup, create-MN, create-concept-brief, design, and structure skills. Trigger: "create note", "create SD", "create spec", "create PIC", "create MN", "create PD", "create FD", "feature definition", "design discussion", "structure outline", "write a report", "executive brief", "create EB", "create campaign", "pulse campaign", or any request to create a vault document by prefix.

Aggregate the day's work into an EOD report and generate tomorrow's SOD (agent context document). Rolls up all closeout entries from the daily note, compares against SOD priorities, runs a mini-retro with insight routing (observation/impact/action/landing zone), discovers goals for user confirmation, then writes the SOD that orient will load tomorrow. On Fridays, also produces an EOW and WRM (Weekly Roadmap). On the last workday of the month, also produces an EOM and MRM (Monthly Roadmap). See the Period Reporting System definition for the full system design. Use this skill when the user says 'end day', 'end of day report', 'EOD report', 'daily report', 'wrap up the day', 'day summary', 'how did today go', 'end-day', or wants a synthesized view of what happened across all sessions. This is different from /closeout (which wraps a single terminal session and creates PICs) -- end-day aggregates ALL closeouts into the day's official record and prepares tomorrow's agent context. Should typically run after the last

Dispatch a team of agents to critically review a spec document (SPC - *.md). Evaluates scope and impact on the actual environment, loads all relevant context (lessons, references, related specs, CLAUDE.md), and flags issues, gotchas, and clarification needs. Use this skill whenever the user wants to review a spec, evaluate a spec before implementation, get critical feedback on a spec, or says anything like "review this spec", "evaluate this spec", "what are the risks in this spec", "is this spec ready". Also trigger when the user passes a SPC file path and wants feedback before planning. This skill focuses on review only — use /create-note PL afterward to produce the plan document and Plane project.

Create or refresh an MRM (Monthly Roadmap) and its WRM (Weekly Roadmap) companion. Audits all open work, aligns it to the user's stated strategic goals, triages PICs into active/parked/blocked, sets monthly objectives with done definitions, derives 3 weekly goals, and produces decision rules for staying focused. See the Period Reporting System definition for the full architecture. Use this skill when the user says "roadmap", "create roadmap", "refresh roadmap", "strategic plan", "what should I focus on", "priority reset", "workload audit", "realign priorities", or after a major meeting changes strategic direction. Also use when the current MRM is expired (past its month) and orient flags "no current MRM."

Contextual skill discovery - observe what the user is working on and surface relevant WFK skills and principles they might not know about. Use this skill when the user says "discover", "what skills should I use", "what can you do", "help me work better", "workflow tips", "what am I missing", or when a new user wants to learn the kit. Can also run as a lightweight tip during /orient.

Use when the user wants to design, redesign, shape, critique, audit, polish, clarify, distill, harden, optimize, adapt, animate, colorize, extract, or otherwise improve a frontend interface. Covers websites, landing pages, dashboards, product UI, app shells, components, forms, settings, onboarding, and empty states. Handles UX review, visual hierarchy, information architecture, cognitive load, accessibility, performance, responsive behavior, theming, anti-patterns, typography, fonts, spacing, layout, alignment, color, motion, micro-interactions, UX copy, error states, edge cases, i18n, and reusable design systems or tokens. Also use for bland designs that need to become bolder or more delightful, loud designs that should become quieter, live browser iteration on UI elements, or ambitious visual effects that should feel technically extraordinary. Not for backend-only or non-UI tasks.

Evaluate incoming materials (files, directories, data) and triage them into the Work Vault's project structure — matching to existing projects or proposing new ones. Use this skill whenever the user points you at files to organize, says "intake these", "file these", "where do these go", "organize these into the vault", "I got files from [someone]", "process these materials", drops a path and says "figure out where this goes", or has external documents (specs, reports, data, reference material) that need a home in the project structure. Also trigger when the user receives deliverables from collaborators (dad, Christian, contractors) and wants them properly filed as reference material for future work. Even casual requests like "dad sent me some stuff, here" or "put these somewhere useful" with a file path should trigger this skill.

Post-implementation code review — reviews changed files, checks for quality issues, generates a review artifact. Use after /implement completes or when the user wants a code review. Trigger on "review the code", "PR review", "code review", "check the implementation", "review what was built", or at the end of an /implement sprint. Do NOT use this for spec review (use /review-spec) or planning (use /create-note PL).

First-run setup for the Workflow Kit vault. Scans the user's machine to auto-discover their work environment, builds a profile, scaffolds the vault with project folders, custom prefixes, and daily note templates. Use this skill when the user says "setup", "set up", "configure the vault", "first time setup", "run setup", or when this is a fresh clone of the workflow-kit repo with no workflow-kit.config.json.

Full-pipeline article intelligence: fetch a web article (or batch of articles), extract clean content, generate summaries, store in the vault, then evaluate against all active projects for relevance and actionable value. If the assessment finds genuine value, dispatches a research team for a deep dive with current best practices, implementation recommendations, and project-specific application notes. Supports single URLs and multiple URLs processed in parallel. Use this skill whenever the user shares an article URL and wants to know if the content is useful for their work, says "check this article", "is this worth reading", "article intel", "analyze this article", "process this article", "what can we use from this", or any variation of wanting web content evaluated against active projects. Also trigger when the user pastes an article link with context suggesting they want evaluation, e.g. "found this on X, might be useful for {{ORG}}" or "someone shared this, looks relevant to the residents portal". Handles s

Break down complex documents, topics, or recent conversation context into clear, contextualized explanations. Use this skill when the user says "explain", "break this down", "what does this mean", "help me understand", "walk me through this", "summarize this spec", "what is this plan saying", or passes a file path to explain. Also trigger when the user seems confused or asks clarifying questions about a complex topic being discussed. Works with any document type (SPC, PIC, PL, RE, ADR, IR, SO, DD) or conversational topics.

Mandatory safety protocol for all git operations — pull, push, merge, commit, checkout, branch creation, and multi-repo coordination. MUST be consulted before ANY git operation that modifies repository state, especially when working across multiple repos, deploying to VPS, dispatching agents to write code, or doing merge/rebase operations. Also trigger when you see dirty working trees, merge conflicts, untracked files blocking operations, worktree agents, parallel git operations across repos, or any git error. This skill prevents the recurring issues of lost work from worktree cleanup, cascading failures from parallel git ops, dirty trees blocking merges, deleted files that shouldn't have been deleted, and branches that never reach remotes. If you are about to run ANY git command beyond `git status` or `git log`, consult this skill first.

Park research, context, or deferred work into a target project's CLAUDE.md so it surfaces when future agents work on that project. Use this skill whenever the user wants to defer context for later use, says "park this", "park context", "defer this to", "queue this for", "save this for later", "save this for when we work on X", "hold this for the portal work", or any variation of wanting to attach context to a project for future pickup. Also trigger on "/park" with arguments. Even casual requests like "we'll need this when we start on X" or "remember this for the portal redesign" should trigger this skill when there's a specific project the context should attach to. This skill is the bridge between "research done now" and "implementation later" — it prevents context from being lost between sessions by embedding it directly in the project's agent config.

Take a reviewed spec (SPC - *.md) and produce a full implementation plan document (PL - *.md), then sync it to Plane as a project with issues, modules, labels, and work items. Reads review artifacts from /review-spec if they exist (scope analysis, context brief, clarification log). Use this skill when the user wants to create a plan from a spec, turn a spec into actionable work items, create a Plane project from a spec, or says "plan this", "create a plan for this spec", "set up the Plane project", "make this actionable". Also trigger when the user has just finished /review-spec and wants to proceed to planning. This is the second half of the review-then-plan workflow — run /review-spec first if the spec hasn't been reviewed yet.

Autonomous sprint mode — user is stepping away and wants work to continue without them. Replaces all user test gates with QA auditor verification, removes soak periods, and sets up Telegram escalation for blockers. Use this skill whenever the user says "step away", "stepping away", "I'm leaving", "run without me", "autonomous mode", "don't wait for me", "keep going without me", "I'll check later", "don't stop working", or any indication they want to leave while an active /implement sprint continues. Also trigger when the user says /step-away. This skill ONLY works during an active implementation sprint with a team (pm-tracker, qa-auditor, workers) — if no team is running, tell the user to start /implement first. Detection: look for any active team containing pm-tracker and qa-auditor roles, regardless of how the team was created (direct /implement or via the implement router's sub-skills like implement-setup).

Systematic diagnostic protocol for when you're stuck on ANY issue — VPS, Docker, Traefik routing, database schemas, frontend/UI rendering, backend APIs, authentication, deployment, LLM pipelines, or third-party integrations. MUST be invoked after 2 consecutive failed fix attempts on the same issue — do NOT wait for the user to ask. Also invoke when the user says 'troubleshoot', 'you're stuck', 'you're going in circles', 'why isn't this working', 'try again properly', or expresses frustration with repeated failures. This skill dispatches a diagnostic agent that verifies assumptions, consults lessons learned, and gathers evidence before any more fixes are attempted. Use this BEFORE trying another fix.

Full-pipeline YouTube video intelligence: transcribe a video or playlist, generate summaries, store in the vault, then evaluate content against all active projects for relevance and actionable value. If the initial assessment finds potential value, dispatches a research team for a comprehensive deep dive with current best practices, implementation recommendations, and project-specific application notes. Supports single video URLs and playlist URLs — for playlists, enumerates all videos, skips already-processed ones, and batch-processes the rest with a summary report. Use this skill whenever the user shares a YouTube URL (video or playlist) and wants to know if the content is useful for their work, says "check this video out", "is this relevant to what we're working on", "research this video", "video intel", "analyze this video", "what can we use from this", "process this playlist", "scan this playlist", or any variation of wanting YouTube content evaluated against active projects. Also trigger when the user p

Scrape website styles into reusable themes, preview themes on the {{ORG}} site via Playwright screenshots, and manage the theme registry. Use this skill whenever the user mentions scraping styles from a website, capturing a site's aesthetic, theme preview, mockups, "what themes do we have", switching themes, "apply this theme", "scrape the theme from [url]", "show me the [name] theme", "preview [name]", or any request about {{ORG}} site theming. Also trigger on /web-theme, "new theme from [url]", "capture the look of [site]", "style guide from [url]", "theme list", or when the user pastes a URL and wants to extract its visual design for {{ORG}}. Even casual requests like "make our site look like [url]" or "I like how [site] looks" should trigger this skill.

Recap the current conversation — surface the main topic, identify side-threads, unanswered questions, unresolved decisions, and loose ends, then triage them interactively. Use this skill when the user says "recap", "where are we", "what were we doing", "what's outstanding", "any loose ends", "what did we miss", "back to the main thing", or when the conversation has branched into multiple topics and needs re-grounding. Also trigger after context compaction when the user seems disoriented, or when the user says "ok what now" mid-session and the conversation has been long or meandering. This skill is the "gather yourself" moment before continuing — it prevents lost context, dropped questions, and forgotten side-tasks.

Set iTerm2 tab title and color to reflect the current work context. Only use this skill when the user explicitly asks: "set tab", "color tab", "label tab", "iterm", or "tag this". Never trigger proactively.

Ingest a spec and produce a Prototype Design Discussion (DD) optimized for Claude Design (claude.ai/design). Extracts user flows, screen inventory, component inventory, data model summary, and interaction patterns from the spec into a single document that can be fed directly to Claude Design for UI prototyping. Use when the user says "prototype", "prototype this spec", "design doc for prototyping", "claude design prep", or wants to create mockups from a spec.

Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, artifacts, posters, or applications (examples include websites, landing pages, dashboards, React components, HTML/CSS layouts, or when styling/beautifying any web UI). Generates creative, polished code and UI design that avoids generic AI aesthetics.

Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.

Analyze Claude Code token consumption from JSONL session logs. Use this skill when the user says "usage audit", "token usage", "what's eating my tokens", "rate limit", "usage analytics", "check my usage", "token consumption", "why am I hitting limits", "full audit", "usage report", "how much am I using", "compare usage", "top consumers", "show me anomalies", "cache efficiency", "which projects use the most", "daily trend", "budget window", or any question about Claude Code token/usage patterns.

Fetch Limitless lifelog data, store it in the vault, and generate meeting notes. Use this skill whenever the user mentions lifelogs, Limitless recordings, "process today's meetings", "what meetings did I have", "update lifelogs", "create meeting notes from recordings", "process yesterday's calls", or any reference to turning Limitless pendant recordings into structured notes. Also trigger on "fetch lifelogs", "sync limitless", "meeting summary from limitless", or when the user asks about meetings from a specific date and lifelogs are the source. Even casual requests like "what did I talk about today" or "summarize my meetings" should trigger this if the user has a Limitless pendant.