Skills directory · page 44

OverKill Hill P³ repository organizer for content-first Git repositories. Use when a local Git repository grew organically from a folder and needs an evidence-based profile, purpose summary, classification, cross-platform naming cleanup, folder design, governance scaffolding, or a safe reorganization plan. Also activate when a repository contains prompts, AI conversations, research, Word documents, spreadsheets, PDFs, images, or mixed knowledge assets rather than a conventional application. This is the authoritative workflow for understanding and organizing an existing repository; use it instead of application-template advice, skill cataloging, or new-repository creation workflows.

OverKill Hill P³ visual style registry and application workflow. Use when a user wants to extract a reusable brand, visual style, palette, typography system, layout language, or interface aura from a website, stylesheet, document, presentation, infographic, or set of visual assets. Also activate when a user wants an SPA, documentation, presentation, image brief, or other target artifact aligned to one or more named style profiles. This is the authoritative profile-driven workflow for capturing and applying visual style without blending unrelated brands or treating inferred cues as declared rules.

Extract pasted or uploaded AI chat threads into standalone, actionable Markdown. Use when the user wants to evacuate context from ChatGPT, Claude, Gemini, Perplexity, Microsoft Copilot, Grok, Mistral Vibe, an unknown AI platform, or a mixed transcript; preserve goals, decisions, reasoning, next actions, prompts, code, files, images, citations, canvases, artifacts, diagrams, and provenance; resolve uncertain speaker boundaries; or requests rapid, balanced, comprehensive, essential, substantial, exhaustive, scan, distill, or catalog depth. Use for complete, partial, flattened, exported, or turn-by-turn human-supplied captures. Do not use for logging into or scraping source services, lossless transcript archival, ordinary short-form summaries, or platform support questions that do not require a durable handoff artifact.

Run a WWGD (What Would Glee Do, Say, Think, Feel) reception check on a draft email, post, article, or message using the Glee-fully persona lens. Scores warmth, kindness, clarity, joy, and "does this sound like the author or like a jerk", then returns keep-or-soften notes. Use when asked for a WWGD check, a warmth pass, or "how will this land". Not a proxy for any real person's opinion.

OverKill Hill P³ client-only Google Identity Services (GIS) auth workflow. Use when designing, implementing, or troubleshooting client-only Google Identity Services (GIS) OAuth token flows for React or TypeScript SPAs on static hosts. Use for Google Sign-In, Calendar or Tasks read access, popup token acquisition, session-scoped token storage, expiry handling, silent re-authentication, GCP OAuth setup, or porting this pattern without a backend, Client Secret, redirect URI, or refresh token. Keep provider-specific write APIs outside this skill.

Plan conversion of an existing Custom GPT into a portable Agent Skill. Use when a GPT already has instructions, knowledge files, actions, conversation starters, test prompts, or usage history and the user needs a capability map, semantic-loss analysis, reusable skill boundary, migration plan, or handoff to okhp3-skill-foundry.

Decide whether an active exploratory effort is ready to become formal, should remain active, should archive, or is blocked. Use when an effort feels ready, the owner asks for a go or no-go, or a promotion package needs a readiness decision. Score evidence rather than enthusiasm and do not create the destination.

Guide a consented still-photo session through verified camera or desktop tools with readiness cues, bounded posing time, stop handling, and saved-file verification. Use for headshot source capture or short posed photo sequences. Distinguishes installed software from usable controls and sequential stills from a true burst. Does not authorize video, device installation, or photo editing.

Assemble all validated BP-SKILL process artifacts into a publication-ready bundle with a manifest, approvals record, and completeness check. Use this skill when the user is ready to publish or hand off a complete process documentation set; when they ask to package the process docs, create a handoff bundle, publish the process artifacts, or generate a release package. Requires a passing validation report (Band A or B) before bundling. Produces a MANIFEST.yaml, APPROVALS.yaml, and a ZIP-ready directory structure.

Validate and prepare a localized static-web page set for release after its exact-pair translation is reviewed. Use when checking locale URLs, BCP-47 html lang values, self-canonicals, hreflang alternates, noindex staging, or rendered-language evidence. It does not translate prose, publish a site, or replace source-drift detection.

Detect which pages on a static site are missing a translation, or have a translation that has fallen behind a changed English source, by reusing the site's own generated search index as the page inventory. Reports drift and names the exact-pair okhp3-translation-en-us-<pair> skill that owns each flagged route. Never drafts, edits, or publishes a translated page itself, and never fails a build over a page the site has not declared in scope for translation.

Turn one raw idea into a short recoverable label without developing it. Use for "park this idea", "capture now, clarify later", or a voice-note idea that should stop interrupting current work. Return a label in conversation; durable capture requires a separately authorized destination and receipt. Do not use for backlog grooming, brainstorming, or executing the idea.

Draft an Instagram comment or reply from supplied post and conversation context. Use when the user wants a context-aware response under a named Instagram post or comment thread. It returns a reviewable draft only and does not post, react, follow, message, moderate, or change comment controls.

Draft an Instagram feed, carousel, Reel, or Story publishing brief from supplied material. Use when the user needs a caption, asset handoff, and visibility or disclosure checklist for a named Instagram post form. It prepares text only and does not create media, set controls, or publish.

Map a specific interview role to a person's supplied career evidence and choose defensible stories. Use for interview positioning, qualification gaps, transferable platform experience, or connecting public work to role needs. Produces preparation material, not applicant rankings or application submission.

Draft candid interview questions that test role fit without loading the premise. Use to reframe an accusatory or overly deferential question, explore executive backing and investment, or ask about a leader's move and priorities. Preserves the concern; does not fact-check a whole company or contact anyone.

Complete or resume job applications from approved career evidence, tailored materials and saved portal drafts. Use for preparing, uploading, submitting, and reconciling application receipts within the user's authorization. Excludes status-only audits, saved-job ranking and recruiter outreach.

Draft a Ko-fi post for a supplied audience and post form. Use when the user needs a public or supporter-access update, blog, image, audio, or embedded video post with an explicit access boundary. It prepares text only and does not inspect eligibility, set audience, save, schedule, or publish.

Draft a private, privacy-preserving Ko-fi reply to a supplied supporter message. Use when the user wants to acknowledge support, answer a simple question, or provide a bounded update after a payment-linked interaction. It returns draft text only and never reads, sends, or manages messages.

Detect abnormal agent-to-agent, service-to-service, and tool-call paths. Use when monitoring authorization boundaries, capability drift, or possible lateral movement. Do not perform active pivots or infer compromise from one event.

Compile a private lifetime career reference or super master from an existing career claim ledger. Use for a detailed career biography, evidence bibliography, experience registers and lessons learned. Source ingestion and claim resolution remain upstream; job-specific resumes are separate.

Mine finished work, repository history, or a current conversation for 3 to 5 evidence-linked LinkedIn angles. Use for "mine this for a post", public progress updates, launch or checkpoint recaps, "what's postable here", or "angles from this". It does not draft or publish the post; use okhp3-linkedin-post or okhp3-linkedin-comment next.

Draft a thoughtful LinkedIn comment or reply to an existing post. Use when the user wants to respond on LinkedIn, contribute to a professional conversation, answer a LinkedIn comment, or add a source-backed public progress, launch, or checkpoint update. It preserves the target context, applies the LinkedIn voice and public-context scrub, and returns a draft only. Do not use for a new standalone post or for publication.

Draft one personalized LinkedIn connection request note per recipient, within the sender's character limit and using only facts the sender supplied. Use when the user wants an invite note, a batch of notes for a planned outreach wave, or a rewrite of a generic connection request. It returns drafts and a gate report only. Do not use for direct messages, replies, or sending.

Check a LinkedIn inbox for new replies from a supplied outreach list, classify each sender, and propose draft counter replies in a fixed report, without sending. Use for a daily or on-demand inbox review after an outreach wave, including unattended scheduled runs. It is read-and-draft only. Do not use to send, mark, delete, or archive messages, or to draft a first-touch message.

Draft a LinkedIn direct message or a counter reply to an inbound message from raw dictation, notes, or the received text. Use when the user wants to write a DM, reconnect after a long gap, answer a new connection's message, follow up after an accepted invite, reply to a recruiter or peer, or coordinate several related messages about one subject. It tiers the recipient by role and relationship temperature, verifies recipient facts, matches length, keeps one small ask, scrubs risk, and returns drafts only. Do not use for comments, posts, connection notes, or sending.

Plan a LinkedIn network expansion: audit followed-but-unconnected accounts, segment and tier targets toward the sender's chosen audience, dedupe against pending and excluded lists, pace outreach in waves, and adjust from acceptance results. Use when the user wants to rebalance or grow a network on purpose. It returns a plan and a wave list only. Do not use for sending, note writing, or message writing.

Draft a source-backed LinkedIn post from a chosen angle or named topic. Use for standalone posts, public progress updates, launch or checkpoint recaps, deadline updates, or "write a post about X". It applies voice and the final public-context scrub, then returns a draft only.

Apply the platform-specific voice contract to any LinkedIn-bound text. Use as the final pass on a post, comment, checkpoint update, or public progress draft, whether produced by another skill or written by hand. It checks prose density, uncertainty, supplied links, hashtags, and public-context handoff. This is a filter, not a generator.

System and solution architecture diagrams in Mermaid for technical audiences - C4 model (Context/Container/Component/Code), architecture-beta cloud/infrastructure diagrams, block diagrams, packet/protocol diagrams, service topology, and integration flows. Use when the user wants to diagram software architecture, infrastructure, deployments, service relationships, or "how systems connect." Always load okhp3-mermaid-core first for audience/type/theming.

BPMN-informed business process modeling in Mermaid. Use whenever the user wants to diagram a business process, workflow, approval chain, decision/gateway logic, swim lanes, cross-department handoffs, onboarding flows, procurement flows, or anything describable as "who does what, in what order, with what decision points." This is the differentiator no community Mermaid skill covers - BPMN vocabulary (gateways, events, tasks, swim lanes, subprocesses) does not exist in mgranberry, WH-2099, softaworks, or Agents365's skills. Always load okhp3-mermaid-core first for audience/type/theming, then this skill for BPMN vocabulary and patterns.

Foundation skill for ALL Mermaid diagram work. Load this first for any task involving Mermaid syntax, diagrams, flowcharts, process maps, architecture sketches, or visualizing relationships/workflows/structures. Handles audience declaration, diagram type selection across all 27 current Mermaid types (stable and beta), the OKHP3 design system, file naming and the diagram registry, and the three-gate validation framework. Routes to okhp3-mermaid-bpmn, okhp3-mermaid-architecture, or okhp3-mermaid-data for domain-specific vocabulary, and to okhp3-mermaid-publish for rendering/output.

Data model and relationship diagrams in Mermaid - entity-relationship (ER) diagrams, class diagrams, schema documentation. Use when the user wants to diagram a database schema, data model, object structure, class hierarchy, or entity relationships with cardinality. Always load okhp3-mermaid-core first for audience/type/theming.

OverKill Hill P³ Mermaid governance profile manager. Use when establishing a visual and behavioral standard for a diagram project, checking a diagram against a declared profile, or auditing multiple diagrams for cross-project coherence. Also activate for GOVERNANCE.md, profile drift, cross-diagram consistency, or a diagram that “doesn’t match the others.” Load okhp3-mermaid-core first. For a palette on one new diagram without a project standard, use okhp3-mermaid-theme-builder instead.

Rendering, exporting, and publishing finished Mermaid diagrams. Use after a diagram has passed okhp3-mermaid-core's three validation gates and needs to become a viewable artifact - local PNG/SVG render, embedding in a .md file, or publishing via the Mermaid Chart MCP for a shareable link. NEVER deletes rendered output; the render IS the deliverable.

Syntax repair for broken Mermaid diagrams. Use when a .mmd file or fenced Mermaid block fails to parse - mmdc reports errors, Mermaid Live shows red, or the rendered output is visually malformed. Diagnoses the parse failure, applies the minimum fix, and re-runs Gate 1 (syntax validation). Does not restructure, restyle, relabel, or redesign. Does not change content. Load okhp3-mermaid-core first. For intentional content changes to a working diagram, use okhp3-mermaid-update instead.

Apply reusable color palettes and visual governance to Mermaid diagram code. Use this skill when the user wants to style, theme, color, or brand a Mermaid diagram; when they want a themeVariables block or %%{init}%% configuration; when they ask for a prompt scaffold that enforces consistent diagram styling for future AI-generated Mermaid; when they want renderer-safe output for GitHub, GitLab, Obsidian, Notion, Confluence, or the Mermaid CLI; when they want to make Mermaid output renderer-safe or renderer-compatible; when they mention Mermaid colors, palettes, CSS variables, diagram cleanup, or on-brand diagrams; or when they want to extract and re-theme existing styled Mermaid code.

Style-preserving update of an existing Mermaid diagram. Use when the user provides an existing .mmd file (or fenced Mermaid block) and a change request - new nodes, revised labels, added edges, restructured flow, or changed content - and the goal is to apply the minimum diff required without touching the diagram's style configuration, classDef declarations, or init block. Load okhp3-mermaid-core first for audience/type context. Do NOT use this skill for syntax repair (broken parse) - use okhp3-mermaid-repair for that.

Detect meaningful changes in approved model behavior, tool use, refusal patterns, or output risk. Use when evaluating prompt injection, poisoning, misconfiguration, or model drift. Do not treat anomaly as proof of malicious intent.

Stress-test a draft article, post, essay, or position piece before publication and return a verdict, load-bearing claims table, weakest point, keep-or-fix judgments on rough edges, and a salvage list. Use when asked to red-team, pressure-test, tear down, or "put it through the furnace". Do not use for security threat modeling or final voice polishing.

Forge and stress-test the opening line and closing line of a short post (LinkedIn, X, Threads, Bluesky, Discord) so the hook earns attention honestly and the close lands hard on the point. Returns scored hook and close options with a pick. Use when a short draft exists and the owner asks to sharpen the hook, fix the ending, or "make it land". Not for full drafting, angle mining, or publishing.

Turn a research packet, rough idea, or brain-dump into a defensible one-sentence thesis, load-bearing outline, hook options, counterargument plan, and a hard ending, ready for drafting and a separate voice pass. Use when asked to find the angle, sharpen an argument, build an article skeleton, or "forge" a thesis. Do not use to write final prose or to publish.

Select NFL fantasy players, DFS lineups, salary-cap rosters, waiver priorities, or trade targets under a scoring and budget system. Use when comparing projected points per cost, replacement value, positional scarcity, floor, ceiling, or roster correlation. Load `okhp3-outcome-modeling-core` and `okhp3-outcome-modeling-sports` first; this is fantasy decision support, not sportsbook betting advice.

Decide where a repeatable Notion-related workflow should live: the in-app Notion Agent, a Notion Custom Agent, a Notion Skill page, a database automation, or an external file-based Agent Skill. Use when the user asks whether something should be a Notion agent, a Notion skill, a Custom Agent, or a portable skill, or describes a recurring workflow and is unsure where to build it. Decides only; does not build or run a Custom Agent and does not convert an artifact between formats.

Use this skill whenever an agent needs to export, capture, ingest, summarize, route, deduplicate, or reconcile a conversation or content artifact into Notion. It guides connector-enabled agents through choosing a Notion page versus database destination, resolving a workspace-specific URL or ID, inspecting page content or data-source schema before writing, mapping thread metadata and reusable extracts, creating or updating records safely, and verifying the result. Trigger for ChatGPT, Claude, Perplexity, Copilot, Gemini, PDF, Markdown, transcript, knowledge-base, Notion API, Notion connector, page, database, data source, or thread-export requests, even when the user does not name this skill.

Use this skill whenever an agent needs to export, capture, ingest, summarize, route, deduplicate, or reconcile a conversation or content artifact into Notion. It guides connector-enabled agents through choosing a Notion page versus database destination, resolving a workspace-specific URL or ID, inspecting page content or data-source schema before writing, mapping thread metadata and reusable extracts, creating or updating records safely, and verifying the result. Trigger for ChatGPT, Claude, Perplexity, Copilot, Gemini, PDF, Markdown, transcript, knowledge-base, Notion API, Notion connector, page, database, data source, or thread-export requests, even when the user does not name this skill.

Load first for any Notion task. Negotiates which Notion surface and tools are actually available (hosted MCP, REST API, or the ntn CLI), pins the API version, enforces the mandatory database-to-data-source resolution rule, and applies Notion's error taxonomy before any other Notion skill runs. Use when a task involves Notion pages, databases, data sources, blocks, or the Notion API or MCP in any way. Performs no read, write, or delete of its own; it is never the final answer to a user request.

The only okhp3-notion-* skill authorized to remove Notion content. Archive, trash, restore, or permanently delete pages, blocks, databases, data sources, and rows, with blast-radius estimation and an explicit confirmation gate before any irreversible step. Use when the user asks to delete, trash, archive, remove, or restore a Notion page, block, database, or row. Does not decide what should be deleted (that is workspace hygiene analysis) and does not remove a schema property from a live data source (use a wave-2 schema-migration skill).

File existing Notion pages into verified domain hubs and reconcile their index records without replacing the originals. Use for root-page clutter, misplaced imports, inbox routing, or a scheduled filing pass that must preserve page IDs, source content, links, and history.

Dependency-loaded by other okhp3-notion-* skills. Resolves any Notion reference format (full URL, notion.site URL, bare UUID with or without dashes, collection:// data-source URL, view:// saved-view URL, discussion:// comment-thread URL, or a plain page/database title) to the correct entity type and ID, and disambiguates database versus data source versus page versus block. Not intended to be selected directly from a user prompt; always loaded by okhp3-notion-core or a sibling that needs an ID resolved before it can act.