Menu
SkillsMP has collected 924 skills from mcorbett51090/RavenClaude. Open a skill to review its source and details.
Showing 40 of 924 collected skills.
Route everyday, well-defined work (single-file edits, test-writing, summarization, mechanical refactors, boilerplate) to a cheaper coding agent (Grok or Copilot CLI) instead of doing it in the main Claude session, escalating…
Authors, validates and packages a claude.ai Organization Skill — the intake questions, the trigger and scope exercises, then lint, pack and verify against measured platform constraints. Reach for this when someone wants to publish a Skill to their…
Drafts a reply to an inbound customer email, grounded in the account record and the published help-centre articles rather than in recall. Use when a customer emails about billing, a renewal or an outage, when a reply must cite current policy, or when a…
Screens a supplier invoice against the purchase order and the approval matrix, flagging quantity, price and duplicate-payment discrepancies before the invoice reaches the approver. Use when an invoice arrives for a purchase order, when a payment run is being…
<What it does, in one clause.> Use when <the concrete situation from your trigger list>, or when <a second situation>.
Write a host-paired handoff and start a fresh window before auto-compact — Claude Code, Grok TUI, Copilot Chat new session, or Copilot CLI. Use for /handoff, 'context is full', 'fresh window', or a Stop-hook hot-window nag.
Bring a RavenClaude install up to date when it's stale, when new skills/agents aren't showing, or after upstream changes ship. Covers both hosts: under GitHub Copilot CLI the update is `git pull` in the marketplace clone then re-running `ravenclaude setup`…
The FORGE gated-planning pipeline that /forge runs: depth-scaled gates that turn any idea into a two-panel-reviewed, critic-checked, tiebroken, routed plan. Mimics Claude Code Ultraplan's deep-plan loop and improves it with cross-model divergence, a…
Team Lead dispatch playbook. Given a feature or task, decide which specialized agents to dispatch, prepare their briefs, allocate worktrees, run them in the right order, and re-route on blockers. Load this skill whenever you (the Team Lead) are about to…
Design the intake behind a form so triage is deterministic: request taxonomy, the fields each request type actually needs, routing rules, per-type response clocks, self-serve-vs-escalate bright lines, and abandonment read as a process defect stream. Produces…
Define a form's measurement contract before instrumenting it: which events, which denominator, what counts as a defect, how per-field drop-off lies, and how to hand an individuals series to statistical process control without manufacturing false signals on a…
Walk the trust boundary of a form submission on the server: client/server validation parity, honeypot design and its assistive-tech exemption, double-submit and submission idempotency, webhook signature verification, PII minimisation. Cites ravenclaude-core…
Make the neutral forms guidance executable on the RavenPower stack: the ordered authority chain for a public write, where the challenge check sits, which fail directions are chosen, and the four honest gaps a reader must re-verify before relying on them.
Design and audit conversion-focused screens — funnel definition, friction-vs-trust balance, form field selection (every field is a cost), one CTA per screen discipline, social-proof placement, microcopy patterns, and the conversion-rate measurement plan.…
Remove finished agent worktrees, prune their branches, and surface anything still in flight. Run at the end of a multi-agent session and weekly as hygiene.
Substrate-neutral pre-execution classifier contract. A single Haiku call emits a `run_config` JSON envelope that right-sizes cardinality knobs + per-phase model tier + reasoning level for multi-phase agentic workflows (the rc-deep-research loop is the first…
Universal pre-dispatch model right-sizer. A single Haiku 4.5 forced-tool call evaluates `{subagent_type, description, prompt_head}` at every Agent dispatch / Workflow `agent()` call / `thing-decide.py` seat dispatch and binds (downgrades AND upgrades) the…
Deep research harness — fan-out web searches, fetch sources, adversarially verify claims, synthesize a cited report. Includes an inline substrate adapter that reads .ravenclaude/run-config.json once at startup; when enabled:false (the default) all agent()…
Create an isolated git worktree under .claude/worktrees/ for a sub-agent to work in. Use this before dispatching any coder agent so that parallel work cannot collide.
Deterministic sanitizer floor for WebFetch return envelopes — strip injection-shaped blocks before any agent treats fetched-body content as authoritative. Two confirmed-in-wild observations (2026-06-02) at ibcs.com/standards and the FT chart-doctor GitHub…
Walk the graph-vs-relational decision tree and return a one-screen verdict — LPG, RDF, stay relational, stay RAG, or stay KV — plus the neighbor hand-off. Reach for this before any labels or Cypher.
Design a retrieval-graph (GraphRAG) extract-index-search architecture only after a BM25/no-graph baseline has lost. Hands paradigm III.a to memory-engineering and eval to ai-rag-engineering.
Design a labeled property graph — labels, typed directed relationships, business-key identity, temporal edges, and a supernode plan. Fill templates/graph-data-model.md before any CREATE.
Write a bounded, typed traversal. Cypher examples first, then ISO GQL, then Gremlin or SPARQL 1.1 stubs. Flag unbounded * and untyped -[]- as bugs. Includes GraphRAG local/global/DRIFT query patterns.
Point at a website home page and extract every logo variant plus the brand 'schema' (design tokens — colors, typography, radii) into a ready-to-apply brand kit, so HTML documents/reports you generate for that project match the brand. Use when the user says…
Capture a reference website's full design schema (spacing/type/grid/elevation/component recipes) and apply it to a target — cloning the craft while swapping in the target's own brand and structurally blocking the reference's identity from leaking. Fidelity is…
Onboards a non-Claude-Code coding agent (GitHub Copilot CLI / OpenAI Codex CLI / Cursor / Aider / Devin Desktop) to this repo at session start. Use when this repo will be operated on by an agent that does not natively read CLAUDE.md. Routes through AGENTS.md,…
Run a periodic staleness sweep over all `plugins/<plugin>/knowledge/*.md` files and any decision-tree sections — flag entries past their `last-verified` window, categorize by re-verification effort (Tier 1-5 per Researcher schema), produce a remediation queue…
Render → see → critique → edit → re-render: the discipline (and a deterministic referee) that lets an agent inspect its OWN rendered output — a web page, a dashboard, a Power BI / Tableau report — and iterate toward correctness/pixel-perfection against…
Open a pull request for the current branch using the project's standard template. Verifies the branch is green, summarizes the diff against main, and pushes only after the user confirms.
Score a consumer repo's GitHub-development protocol against the shipped gold-standard catalog and produce a remediation queue ranked by leverage. A 10-row core rubric (mapping 1:1 to the P3 exemplar-repo catalog + the P4 Actions-hardening rules) plus 3…
Author a Vega-Lite / Deneb / SVG spec for a stated intent on any surface (web vega-embed, react-vega, Evidence, Observable, Power BI Deneb, Tableau extension/SVG, SVG-in-DAX). Six-step method: pick grammar → bind data → encode → wire interactivity → test…
Deterministic layout-arithmetic linter for dashboard/report page definitions. Runs seven checks — no-overlap (AABB), within-canvas, equal-gap, column-alignment, and three PBIR-specific invariants (no-empty-binding, bounded theme-override count,…
Lint a standalone SVG file for geometry soundness, legibility, and security before committing to a report or embedding in Power BI/Tableau. Checks viewBox presence and aspect ratio, minimum text font-size, and the security floor: no <script>, no on* handlers,…
Two fresh-independent expert panels review a strategic plan, fill its gaps, author a tactical build plan, then a different panel cold-reviews the build plan and emits P0/P1 recommendations. Includes an upfront advisory routing analysis recommending Ultraplan…
Safely retire a local branch by tagging its tip first, then deleting the branch. The sanctioned escape hatch from guard-destructive.sh's blanket `git branch -D` block — turns silent-loss-of-unmerged-work into a reversible, audit-logged operation. Use when an…
Put a defensible number on a memory system — break-even against a named baseline, growth and caps, cache-invalidation cost, and cost per correct answer. Reach for this on any pay-for-itself question.
Prove a memory system earns its write path — golden set with provenance, judged failure modes, the runnable bake-off, and cost per correct answer. Reach for this before adopting, replacing or retiring any memory system.
Decide whether to build a memory system at all, and if so which paradigm — raw context, flat retrieval, LLM-extraction or agentic — behind a mandatory baseline-first gate. Reach for this before any memory design work.
Design retention and erasure before the first write — TTL, caps, decay, consolidation timing, dedup, contradiction handling, and what a delete actually leaves behind. Reach for this on any durable store.