beam-ai-team
170 verified skills
langfuse-connect
Connect and operate Langfuse observability workflows for traces, observations, scores, prompts, datasets, queues, and model metadata. Use when the user asks for Langfuse setup or trace analysis.
data-ai
tavily-gtm-research
Load when user mentions "tavily research", "market intelligence", "competitive research", "GTM research", or needs real-time market data for sales, marketing, or vertical strategy.
data-ai
notion-connect
Connect to any Notion database by name. Load when user mentions 'notion', 'connect notion', 'setup notion', 'query [database-name]', 'add to [database]', 'notion databases', or any database name from persistent context. Meta-skill that discovers workspace, caches schemas, and routes to appropriate operations.
data-ai
workable-master
All Workable ATS operations — fetch JDs, search candidates, post assessments/reviews. Load when user says "fetch JD", "search workable", "push to workable", "post review", "rate candidate", "workable", "push assessment", "list jobs", or after interview-coach completes an evaluation. Replaces workable-fetch-jd and workable-push-assessment.
tools
notebooklm-connect
notebooklm, create notebook, audio overview, podcast.
content-media
notebooklm-master
Shared resource library for NotebookLM Enterprise integration skills. DO NOT load directly - provides common references (setup, API docs, error handling, authentication) and scripts used by notebooklm-connect and individual NotebookLM skills.
development
miro-connect
Connect to Miro for board management, sticky notes, cards, shapes, frames, connectors, and collaboration. Load when user mentions 'miro', 'miro board', 'sticky note', 'whiteboard', or any Miro operations. Meta-skill that validates config and routes to appropriate operations.
testing
get-beam-agent-graph
Fetch and save Beam.ai agent graph JSON via API. Load when user says "get agent graph", "fetch beam graph", "download agent graph", "beam agent graph", "get graph for agent", or provides a workspace ID and agent ID to retrieve the graph structure.
development
archive-project
Load when user says "archive project", "archive [project-name]", "move to archived". Moves completed projects to 05-archived/ folder for clean project list.
tools
beam-master
Shared resource library for Beam AI integration skills. DO NOT load directly - provides common references (setup, API docs, error handling, authentication) and scripts used by beam-connect and individual Beam skills.
development
1on1-review
Deep expert coaching review of a 1-on-1 meeting. Triggers: '1on1 review', 'review my 1on1 with [name]', 'coach me on my 1on1', 'how did my 1on1 go', 'deep review my 1on1', 'expert feedback on my meeting with [name]'.
development
blackbird
Weekly reflection that helps you see yourself. Load when user says "blackbird", "reflect", "review my week", "coach me", "am I on track", or "what should I focus on".
tools
create-meeting-minutes
Transform a meeting transcript into formatted meeting minutes. Load when user says "create meeting minutes", "format meeting notes", "generate minutes from transcript", "meeting minutes from recording", or provides a transcript to format.
development
create-skill
create skill, new skill, automate this, make reusable.
tools
candidate-compare
Compare multiple candidates for a role and recommend the best fit. Load when user says "compare candidates", "select best candidate", "who should we hire", "compare [names] for [role]", "rank candidates", or "best candidate for [role]".
tools
1on1-prep
Prepare for an upcoming 1-on-1. Triggers: 'prep my 1on1 with [name]', '1on1 prep', 'prepare for my meeting with [name]', 'what should I cover with [name]', 'get ready for 1on1'.
tools
dealroom-creation
Quick company folder setup for sales prospects. Load when user mentions "create company", "new company", "create dealroom", "company folder", or "setup [company name]". Creates simple folder structure for organizing company documents, call transcripts, and emails.
documentation
mental-models
mental model, think through this, help me decide, first principles, pre-mortem.
tools
expert-improve
Iterative expert refinement until work is excellent. Load when user says: 'improve this', 'make this better', 'polish until ready', 'iterate with experts', 'refine this'
tools
lead-research
Research sales leads and assess fit for Beam. Use when preparing for prospect calls, evaluating inbound leads, or qualifying opportunities. Keywords: lead research, call prep, prospect research, fit assessment, sales call preparation, qualify lead, discovery call prep.
business
prompt-versioning
Manage AI prompts with version control, rollback, and Linear integration. Load when user mentions 'prompt versioning', 'version prompt', 'rollback prompt', 'prompt history', 'create prompt version', or needs to manage prompts across client projects.
tools
qa-checklist
Transform technical agent/node documentation into business-owner cheatsheets. Use when asked to create "cheatsheet", "review guide", "QA checklist", "validation guide", or when business owners need to verify AI decisions. Applies better-doc principles. Outputs outcome-focused documents without technical/node details.
testing
google-docs
Read, write, and manage Google Docs. Load when user mentions 'google docs', 'google document', 'create doc', 'read doc', 'write doc', 'edit document', or references creating/editing text documents in Google Drive.
development
use-case-proposal
Build a Palantir-shape, PDF-native use-case proposal document for a sophisticated enterprise account: research-grounded use cases (each with description, challenge, impact, value), an operating-graph ontology page, a recommended PoC with a week-by-week plan, and a closing page that asks for one decision. Load when a client asks us to 'propose high-impact use cases', requests a use-case presentation/catalog for a function (finance, HR, ops), or when a technical evaluation team will review candidates to pick a PoC. NOT for single-account cold outreach (use prospect-brief), full process diagnostics (use operating-diagnostic), or priced proposals (use proposal-creation).
tools
google-sheets
Read and write Google Sheets data. Load when user mentions 'google sheets', 'spreadsheet', 'update sheet', 'read sheet', 'append to sheet', or references extracting data to update a tracking sheet.
development
google-speech-to-text
Transcribe audio files to text using Google Cloud Speech-to-Text API. Load when user mentions 'transcribe', 'speech to text', 'audio to text', 'transcribe audio', 'voice to text', 'transcription', or converting audio/recordings to text.
development
google-calendar
Read, create, and manage Google Calendar events. Load when user mentions 'google calendar', 'calendar', 'schedule', 'meeting', 'event', 'appointment', 'book time', 'check availability', 'find slots', 'free time', or references scheduling/calendar operations.
development
happenstance-connect
Search and research LinkedIn connections via Happenstance API. Use when user says: 'search my network', 'find someone on LinkedIn', 'research a person', 'who do I know that...', 'happenstance search', 'look up a contact'.
development
hubspot-master
Shared resource library for HubSpot integration skills. DO NOT load directly - provides common references (setup, API docs, error handling, authentication) and scripts used by hubspot-connect and individual HubSpot skills.
development
langfuse-admin
Langfuse project and organization administration. Load when user says 'create project', 'delete project', 'api keys', 'org members', 'list projects', 'update project', 'manage members'.
development
figma-connect
Connect to any Figma file for data extraction and asset export. Load when user mentions 'connect figma', 'get figma file', 'figma URL', 'export from figma', 'figma assets', or provides a Figma file URL.
data-ai
langfuse-models
Manage Langfuse model definitions for cost tracking. Load when user says 'list models', 'create model', 'get model', 'delete model', 'model costs'.
data-ai
gmail
Read, send, and manage Gmail emails. Load when user mentions 'gmail', 'email', 'send email', 'check email', 'inbox', 'draft', or references sending/reading emails.
testing
google-master
Shared resource library for Google integration skills. DO NOT load directly - provides common references (setup, API docs, error handling, authentication) and scripts used by gmail, google-docs, google-sheets, google-calendar, and future Google skills.
development
how-beam-next-works
Deprecated compatibility alias. Learn the current Prism/Beam mental model with the consolidated learn skill. Use `learn` with mode `overview` instead of this legacy Beam Next learning skill.
data-ai
fathom-fetch-meetings
Fetch meeting recordings and transcripts from Fathom API filtered by attendee email domain. Load when user says "fetch meetings", "get fathom recordings", "fathom meetings for [domain]", "list meetings from fathom", "get transcripts", or mentions Fathom.
development
google-gemini-image
Generate, edit, and refine images with Google Gemini. Load when user says 'generate image', 'edit image', 'refine image', 'text to image', 'gemini image', 'modify image'.
development
update-variables-in-notion
Extract variables from Beam prompt chain files (.md) and sync to Notion database. Load when user mentions "extract variables", "prompt variables to notion", "update variables", "document variables", or provides a path to prompt files. Handles single files or directories. Creates/updates Notion database with variable metadata (name, description, data type, node).
documentation
setup-notion-customer-onboarding
Setup and fill Notion customer onboarding template pages by updating Overview sections, Quick Links, and database tasks. Load when user mentions "setup notion onboarding", "fill customer template", "update onboarding tasks", or provides a Notion onboarding page URL. Uses Notion MCP for all content updates.
tools
beam-design-system
Access Beam's cached design system tokens and components. Load when user mentions 'beam design system', 'beam colors', 'beam typography', 'design tokens', 'brand colors', 'beam fonts', or asks about Beam's design standards.
content-media
create-project
Create new projects. Load when user wants to START something new with a deliverable endpoint. Keywords: create project, new project, start project, plan project.
tools
email-automation
Load when user says "process emails", "triage inbox", "email automation", "check emails", or "inbox zero"
tools
ui-developer
Frontend UI development for Beam Studio V2 with Tailwind and design tokens. Load when user needs Beam Studio UI, components, or Tailwind layout work.
development
document-nodes
Generate human-readable documentation and checklists from Node JSON files. Use when asked to document nodes, create checklists for node processing, or explain node extraction logic. Processes all nodes in the Nodes directory systematically.
testing
create-master-skill
create master skill, integration master, shared skill library.
development
notion-master
Shared resource library for Notion integration skills. DO NOT load directly - this provides common references (setup, API docs, error handling, database schema) and scripts used by notion-connect, query-notion-db, import-skill-to-beam-next, and export-skill-to-notion.
development
slack-master
Shared resource library for Slack integration skills. DO NOT load directly - provides common references (setup, API docs, error handling, authentication) and scripts used by slack-connect and individual Slack skills.
development
validate-skill-functionality
validate skill, check skill, review skill execution.
testing
slack-connect
Connect to Slack workspace for messaging and channel management. Load when user mentions 'slack', 'connect slack', 'slack message', 'slack channel', 'send to slack', or any Slack operations. Meta-skill that validates config, discovers workspace, and routes to appropriate operations.
testing
call-prep
Sales call prep wizard. Researches a prospect, identifies buyer persona and vertical, generates a complete battle plan. Load when user mentions 'prep', 'call prep', 'prep for [company]', 'meeting prep', 'sales prep', 'prepare for [name]', 'demo prep', or any upcoming sales/demo call preparation.
business
okr-self-review
Generate a structured performance self-review. Load when user says "review my performance - [name]", "review my performance", "self-review", "performance review", or "quarter review". Pulls all data from Notion (OKRs, Key Projects, Skills, Core Behaviours) and cross-references with Nexus chat logs, BEO Slack data, and Fathom meetings.
data-ai
learn-skills
Deprecated compatibility alias. Learn Prism skills and reusable workflows with the consolidated learn skill. Use `learn` with mode `skills` instead of this legacy Beam Next learning skill.
tools
learn-integrations
Deprecated compatibility alias. Learn Prism integrations, connectors, and MCP with the consolidated learn skill. Use `learn` with mode `integrations` instead of this legacy Beam Next learning skill.
tools
langfuse-prompts
Manage Langfuse prompts and prompt versions. Load when user says 'list prompts', 'create prompt', 'get prompt', 'delete prompt', 'update prompt version'.
data-ai
langfuse-scores
Manage Langfuse scores and score configs. Load when user says 'list scores', 'create score', 'delete score', 'score configs', 'evaluation scores', 'add score'.
tools
langfuse-queues
Manage Langfuse annotation queues, items, and assignments. Load when user says 'list queues', 'create queue', 'annotation queues', 'queue items', 'assign reviewer', 'unassign reviewer', 'annotate item'.
tools
langfuse-traces
Query and manage Langfuse traces, observations, and sessions. Load when user says 'list traces', 'get trace', 'delete traces', 'list observations', 'list sessions', 'show spans', 'session details'.
data-ai
amie-connect
Work with Amie calendar and meeting notes. Load when user mentions "amie", "meeting notes", "get transcript", "save transcript to file", "export transcript", "action items from meeting", or wants to access Amie data.
data-ai
langfuse-status
Check Langfuse health and usage metrics. Load when user says 'health check', 'is langfuse up', 'metrics', 'usage stats', 'check status'.
testing
langfuse-datasets
Manage Langfuse datasets, items, runs, and run-items. Load when user says 'list datasets', 'create dataset', 'get dataset', 'dataset items', 'dataset runs', 'add test case', 'evaluation results', 'delete dataset item', 'delete dataset run'.
testing
langfuse-ingestion
Langfuse data ingestion, comments, and media. Load when user says 'batch ingest', 'otel ingest', 'opentelemetry', 'list comments', 'create comment', 'get media', 'upload media', 'bulk import'.
documentation
linear-update-tickets
Update Linear ticket status or add comments. Load when user says "update linear ticket", "mark ticket as done", "move CLI-XXX to [status]", "change ticket status", "add comment to ticket", or "close ticket".
tools
setup-linear-onboarding-template
Interactive assistant to fill Linear onboarding template projects with content from user's workspace. Load when user mentions "setup linear template", "fill linear project", "onboard client in linear", or provides a Linear project URL. Guides user through each task, suggests content from workspace folders, and updates Linear issues with approved content.
tools
setup-goals
Deprecated compatibility alias. Personalize Prism with role, goals, preferences, and useful work context. Use `setup` with mode `personalize` instead of this legacy Beam Next learning skill.
development
learn
Consolidated Prism learning guide. Use when the user asks how Prism or Beam works, wants a tutorial, asks about Projects, Goals, Skills, Agents, Knowledge Base, Integrations, MCP/connectors, or wants a deeper Beam/Prism mental model. Supports sections: overview, projects-goals, skills, integrations, advanced.
tools
setup-workspace
Deprecated compatibility alias. Organize Prism sources, Knowledge Base context, integrations, and workspace setup. Use `setup` with mode `workspace` instead of this legacy Beam Next learning skill.
tools
beam-debug-issue-tasks
Debug failed/issue tasks from Beam.ai using Langfuse traces. Load when user says "debug issue tasks", "check failed tasks", "why did task fail", "task errors", "debug agent", or needs to investigate task failures.
development
beam-graph-creator
Build and deploy complete Beam agent graphs with conditional branching, convergence nodes, and multi-path routing. Takes a specification document and flowchart (builds flowchart on Miro if not provided), designs all nodes/edges/params, deploys via PUT/PATCH, and validates with test data. Use when user says "create beam agent graph", "build beam graph", "deploy agent", "new beam agent", "build graph from spec", "create nodes and edges", or needs to go from a specification document to a working deployed agent on Beam.
development
amie-master
Shared resource library for Amie integration skills. DO NOT load directly - provides common references and scripts used by amie-connect and individual Amie skills.
development
fathom
Complete Fathom integration skill. Load when user wants to fetch meetings, get transcripts, list recordings, or interact with Fathom meeting data. Supports filtering by attendee domain and extracting summaries/action items.
data-ai
google-drive
Manage Google Drive files and folders. Load when user mentions 'google drive', 'drive', 'upload file', 'download file', 'share file', 'create folder', or references cloud file storage operations.
development
reduce-hallucination-archived
Audit a Beam agent for hallucination risk using interrogation-science techniques — check prompts, input/output schemas, node boundaries, and (when available) real task evidence; then recommend prompt fixes, schema fixes, and graph-level additions (Code Executor validators, LLM cross-examiner nodes, abstention routing). STRICTLY READ-ONLY: never modifies, publishes, or deploys anything — output is an audit report with ranked recommendations. Trigger on "audit this agent for hallucination", "why is this agent extracting wrong data", "check this prompt for hallucination risk", "reduce hallucination", "审计幻觉", "为什么提取不准", "检查这个 agent 的幻觉风险", or when a user shows correction/complaint data from a production agent.
development
abstract
Zoom out and contextualize overwhelming concepts. Load when user says 'zoom out', 'abstract', 'I'm lost', 'I'm overwhelmed', 'bigger picture', or expresses confusion about where a concept fits. Identifies paradigms, shows hierarchies, teaches with analogies, and connects to user's goals.
development
impeccable
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.
development
taste-skill
Anti-slop frontend skill for landing pages, portfolios, and redesigns. The agent reads the brief, infers the right design direction, and ships interfaces that do not look templated. Real design systems when applicable, audit-first on redesigns, strict pre-flight check.
development
share-skill
Package and publish a local SKILL.md workflow to the shared skills catalog. Use when the user wants to share, upload, publish, or distribute a skill with the team.
tools
quick-start
Deprecated compatibility alias. Complete Prism onboarding setup with the consolidated setup skill. Use `setup` with mode `quick-start` instead of this legacy Beam Next learning skill.
tools
create-data
Generate synthetic training data from a ruleset file. Use when the user wants to create training examples, generate augmented data, or produce samples following defined rules.
data-ai
beam-put-payload-builder
Build valid PUT payloads for Beam agent graphs by transforming GET responses into Swagger-DTO-compliant format. Handles link translation (linkParamOutputId UUID → linkedOutputParamNodeId + linkedOutputParamName), field remapping, and computed field stripping. Load when user says "build put payload", "put beam graph", "rebuild graph payload", "transform graph", "put payload builder", "restore beam graph", "get to put", or needs to send a PUT to /agent-graphs/{agentId}.
development
calculate-beam-agent-pricing
Design node architecture and calculate comprehensive pricing for Beam AI agents based on requirements. Load when user says 'calculate agent pricing', 'price this agent', 'design agent architecture', 'estimate agent cost', 'node breakdown for agent', or needs detailed cost analysis for a Beam agent project. Generates complete node-by-node breakdown with credit consumption, monthly economics, optimization strategies, and client pricing models.
tools
generate-demo-agent-sample-input-data
Generate sample test inputs and expected results for Beam.ai agents. Load when user says "generate demo data", "create sample inputs for agent", "demo data for beam agent", "test data for agent", "sample input data", or provides a Beam agent URL asking for test cases.
testing
create-beam-task-auto-trigger
Create an auto-trigger webhook script for a Beam.ai agent task. Supports simple query-based, query with attachments/data (base64-encoded), and email-triggered (Gmail/Workable) agents. Load when user says "create auto trigger", "create task auto trigger", "convert task to webhook", "create beam webhook", "automate beam task", or provides a Beam task URL to create a webhook trigger script.
development
demo-documentation-generation-agent
Generate comprehensive, fully-detailed documentation for Beam AI agents from graph.json files. Load when user says "document beam agent", "create agent documentation", "generate agent docs", "document this agent", "demo documentation", or provides a Beam agent graph file. Analyzes graph structure, extracts workflow, tools, integrations, descriptions, and generates complete documentation with NO placeholders.
tools
heyreach-connect
Connect and operate HeyReach LinkedIn outreach workflows. Use when the user asks about HeyReach campaigns, LinkedIn automation, campaign stats, lead import, sender accounts, or campaign operations.
tools
create-meeting-presentation-html
Generate a self-contained HTML slide presentation for meetings. Load when user says "create presentation", "meeting presentation", "html presentation", "slide deck", "make slides", "presentation for meeting", "create slides for". Saves HTML next to session context when possible — account projects (03-projects/accounts), lead folders (04-workspace/leads), or customer workspace — not only 04-workspace root. User describes use case, context, and goals; AI produces a polished dark-themed HTML deck with arrow-key navigation.
development
better-doc
Improve documents for clarity, precision, and impact. Load when user says "better doc", "improve writing", "make clearer", "smart brevity", or provides text asking for style improvements. Uses Classical Style (Pinker) and Smart Brevity (Axios) principles.
testing
execute-project
continue [build], resume, work on, check progress, execute build, build [ID/name].
development
import-skill
import skill, download skill, install skill, get skill.
tools
validate-skill-functionality
validate skill, check skill, review skill execution.
testing
write-integration-ticket
Write API integration tickets in structured specification format. Load when user mentions 'write integration ticket', 'create integration spec', 'integration specification', 'API action ticket', or wants to document an API integration with endpoints, parameters, acceptance criteria, and error handling. Creates Linear tickets with Integration label assigned to Adeel Akram.
development
airtable-connect
airtable, connect airtable, query [base], add to [table].
data-ai
process-client-meeting
Complete end-to-end workflow to process a client meeting. Load when user says "process meeting", "meeting workflow", "process client meeting", "full meeting processing", "handle meeting for [client]", or "post-meeting workflow". Handles transcript fetching, minutes creation, Linear tickets, client updates, internal updates, and project context sync.
tools
create-deal
Create a new HubSpot deal after a successful sales call where a demo has been committed. Use this skill whenever the user says a call went well and a demo was booked, or asks to "log a deal", "create a deal from this call", "add to HubSpot", or "turn this into a deal". Only trigger when there's a clear signal that a demo or next commitment was secured — do NOT trigger for exploratory calls with no outcome, post-demo follow-ups, or existing deals being updated. Pulls notes from Granola automatically, extracts key fields, asks the user for any missing info, and creates the deal in HubSpot with all required fields and associations.
testing
follow-up-automation
Streamlined follow-up system for sales calls based on call transcripts. Load when user mentions "follow-up", "follow up email", "call follow-up", "send follow-up", "post-call follow-up", "follow-up automation", "create follow-up", "generate follow-up", or when user provides a call transcript and needs to send follow-up communication and update CRM. Analyzes call transcripts to extract technical requirements, decision makers, timeline, and pain points, then generates concise technical follow-up emails (under 200 words) and updates CRM with lead status and next steps.
tools
okr-self-review
Generate a structured performance self-review. Load when user says "review my performance - [name]", "review my performance", "self-review", "performance review", or "quarter review". Pulls all data from Notion (OKRs, Key Projects, Skills, Core Behaviours) and cross-references with Nexus chat logs, BEO Slack data, and Fathom meetings.
data-ai
format-and-lint-code
format code, lint code, pre-commit, before commit, code quality, auto-format
tools
evaluate-solutions-case-study
Load when user says 'evaluate case study', 'review candidate submission', 'analyze case study', 'candidate strengths and weaknesses', 'assess candidate work', or 'evaluate solutions engineer candidate'. Systematically evaluate Solutions AI Engineer candidate case study submissions by analyzing extraction prompt quality, project plan depth, and presentation/communication.
testing
beam-ape-optimizer
Automated Prompt Engineering (APE) optimization loop for Beam agent nodes. Adds reasoning traces, runs test batches against a golden dataset, uses 3-agent credit assignment (Doer/Critic/Editor) to identify and fix failing prompt instructions, and redeploys improved prompts. Use when user says "optimize prompts", "APE loop", "improve beam agent accuracy", "prompt engineering", "test and fix prompts", "run APE", "credit assignment", or when agent accuracy needs systematic improvement rather than manual prompt editing.
development
add-integration
add integration, new integration, integrate with, connect to [service].
tools
create-ruleset
Create a ruleset for synthetic data generation. Use when the user wants to define rules for generating training data, create a new dataset domain, or set up data augmentation guidelines.
documentation
git
Git operations for Nexus. Load when user mentions: git, sync upstream, fetch upstream, fetch origin, push origin, pull, commit, git status, git add, git log, git branch, git stash, git reset, git revert, undo commit, update from template, setup git, ssh key, clone repo, git config.
tools
expert-review
Quick expert feedback on any work. Load when user says: 'expert review', 'what do experts think', 'get feedback', 'review this', 'is this good?', 'what's wrong with this?'
development
markdown-to-pdf
Convert markdown files to professionally styled PDFs. Triggers: 'render pdf', 'markdown to pdf', 'create pdf', 'export pdf', 'generate pdf', 'make pdf'.
tools
list-skills
list skills, show skills, what skills, available skills.
data-ai
plan-project
create build, new project, start build, plan build, build [X].
development
search-skill-database
search skills, find skill, browse skills, skill marketplace.
data-ai
google-connect
Connect Google Workspace and route requests across Gmail, Docs, Sheets, Calendar, Drive, Tasks, and Slides. Use when the user needs Google setup, OAuth validation, or integration troubleshooting.
development
create-weekly-update
Generate weekly status update based on Linear ticket activity and project progress. Load when user says "weekly update", "weekly status", "create weekly report", "status update for [project]", "what got done this week". Fetches ticket changes and generates formatted summary.
development
google-tasks
Manage Google Tasks and task lists. Load when user mentions 'google tasks', 'tasks', 'todo list', 'create task', 'complete task', or references task/todo management.
development
generate-linear-project-update
Generate concise Linear project status updates with issues grouped by status, sub-issue links, and metrics. Load when user mentions "generate project update", "weekly update for [project]", "Linear project summary", "status update", or provides a Linear project URL.
testing
langfuse-master
SHARED RESOURCES - DO NOT LOAD DIRECTLY. Contains Langfuse API client, config checker, and reference documentation. Used by langfuse-connect and all operation skills.
tools
linkedin-outbounding-automation
Automate LinkedIn outbound from HubSpot company lists. Load when user says "linkedin outbounding automation", "linkedin outreach from hubspot", "hubspot to heyreach", "prospect C-levels", "run linkedin campaign", or wants to push HubSpot segments to LinkedIn outreach via HeyReach.
tools
linear-create-tickets
Create Linear tickets from action items or task lists. Load when user says "create linear tickets", "add tickets to linear", "create tickets from meeting", "linear ticket for [task]", or wants to convert action items into Linear issues.
tools
ui-developer
Frontend UI development for Beam Studio V2 with Tailwind and design tokens. Load when user needs Beam Studio UI, components, or Tailwind layout work.
development
setup
Consolidated Prism-first onboarding setup. Use when the user asks to get started, quick start, set up Prism, personalize Beam/Prism, define goals, configure their workspace, organize sources, or set up their first useful workflow. Supports modes: quick-start, personalize, workspace.
development
beam-get-agent-analytics
Get analytics for a Beam.ai agent including task counts, completion rates, evaluation scores, and performance trends. Load when user says "agent analytics", "beam analytics", "agent performance", "task stats for agent", "how is agent performing", or needs agent metrics.
data-ai
beam-next-skills-audit
Audit registry.yaml against disk, validate SKILL.md frontmatter, find duplicates and orphans. Load when user says 'audit skills registry', 'validate beam-next-skills', 'registry drift', 'skills catalog audit', 'check registry yaml'.
testing
select-llm-model
Select optimal LLM model for Beam AI tools based on cost, context, and performance. Load when user says "select llm model", "choose model", "optimize llm", "which model for", "model recommendation", "best model for agent", "llm cost optimization", or needs help selecting the right AI model for their Beam tools.
tools
numbers
Use when communicating quantitative information in any form — Slack updates, emails, reports, decks, dashboards, landing pages, product UI, public talks. Covers two integrated layers: (1) making numbers semantically meaningful (translation, anchoring, simplification, story-pairing) and (2) showing numbers cleanly (chart vs table vs prose, chart-by-message, pre-attentive emphasis, color discipline, decluttering). Distilled and integrated from *Show Me the Numbers* (Stephen Few) and *Make Numbers Count* (Chip Heath & Karla Starr). Not for raw data analysis or statistics — this is about communication of numbers, not their derivation.
development
miro-master
Shared resource library for Miro integration skills. DO NOT load directly - provides common references and scripts used by miro-connect.
development
analyze-competitors
Conduct competitor analysis for product design. Load when user says 'analyze competitors', 'competitor research', 'competitive analysis', 'check how [competitor] does this'. Guides exploration of Zapier, n8n, Lindy, Relevance.ai and other automation platforms.
tools
heyreach-master
Internal resource library for HeyReach integration. Contains shared API client, operation scripts, and reference documentation.
tools
analyze-context
Analyze user-provided context for Prism. Use when the user wants to upload, connect, summarize, or organize files, folders, documents, repos, transcripts, screenshots, or other source material so Prism can use it through Knowledge Base / Sources. Prefer Prism-native source ingestion and evidence summaries; only use local Beam Next folders when explicitly requested.
testing
learn-projects
Deprecated compatibility alias. Learn Prism Goals, Projects, Tasks, and planning with the consolidated learn skill. Use `learn` with mode `projects-goals` instead of this legacy Beam Next learning skill.
development
beam-agent-manager
Orchestrate Beam agent graph operations — create, modify, and optimize agent graphs via the right child skill. Covers graph creation (from YAML, from spec, or simple POST), structural changes (PUT), prompt/model/param updates (PATCH), prompt optimization (APE), and publishing. Use when user says "beam agent manager", "create beam agent", "build beam graph", "update beam graph", "patch and publish", "structural change", "add nodes", "optimize prompts", "APE loop", or any graph modification task. Do NOT use for runtime operations (tasks, analytics, HITL, debugging) — those are handled by beam-connect.
development
beam-credit-analysis
Analyze Beam.ai agent credit consumption per execution path. Load when user says 'credit analysis', 'beam credit analysis', 'agent credit consumption', 'how many credits does this agent use', 'cost per path', 'analyze agent credits', 'credit breakdown', or provides a Beam agent URL and asks about credits or cost.
data-ai
beam-graph-edit
Edit Beam agent graph nodes — update prompts, input/output param descriptions, and push changes via PUT. Load when user says 'edit beam graph', 'update beam node', 'change beam prompt', 'push graph changes', 'modify agent graph', 'update param description', or needs to modify a live Beam agent's graph configuration. Uses the proven GET-clean-modify-PUT cycle that bypasses the unreliable PATCH endpoint.
data-ai
build-beam-agent
End-to-end agent builder for Beam AI. Takes a demo request or requirements doc, designs the workflow collaboratively with the user (via mermaid diagrams), then deploys the agent to Beam AI via API. Triggers on "build beam agent", "create beam agent", "deploy agent to beam", "build agent from request", "build demo agent". Handles the full lifecycle from requirements to live agent on the platform.
development
create-beam-agent
Create a Beam AI agent from a YAML description by generating a JSON spec and deploying it via the complete graph API. Load when user says "create beam agent", "deploy beam agent", "build a beam agent", "create agent from yaml", or provides a YAML agent spec and wants it deployed to Beam.
development
beam-connect
Connect to Beam AI workspace for runtime agent operations. Load when user mentions 'beam', 'beam agent', 'beam task', 'beam analytics', 'list agents', 'create task', 'debug task', 'feedback automation', 'agent pricing', or any Beam AI runtime operations. Meta-skill that validates config, discovers agents, and routes to appropriate operations. For build-time operations (create/modify/optimize agent graphs), use beam-agent-manager instead.
tools
1on1-followup
Quick post-meeting follow-up for 1-on-1s. Triggers: '1on1 followup', 'follow up on my 1on1', '1on1 with [name]', 'debrief my 1on1', 'process my 1on1 with [name]'.
tools
bulk-complete
Bulk-complete tasks/steps in a project. Load when user says "bulk complete [project]", "mark all done", "complete phase X", or needs to finish project tasks efficiently. Works with both steps.md and tasks.md formats.
development
beam-teach
Stateful multi-session tutor adapted for Beam — teach a stakeholder to understand, trust, and operate a specific agent, or teach a Solution Engineer a client's business process for delivery. Grounds every lesson in Knowledge Hub sources (real agent graphs, real tasks, transcripts, Linear) before any web resource. Also works for any general topic. Trigger on "teach me", "beam teach", "教我", "onboard <person> on <agent>", "help <stakeholder> understand the agent", "learn this client's process".
tools
create-client-project
Create a new client project folder with complete structure and templates. Load when user says "create client project", "new client", "add client [name]", "onboard new client", "setup client folder", or "create project for [client]".
tools
data-baker
Generate realistic synthetic datasets for AI agent training and testing across diverse domains. Use when user provides meeting transcripts, requirements documents, process flows, or requests synthetic data generation for identity documents, resumes, financial records, medical records, legal contracts, product catalogs, customer profiles, organizational knowledge, or email communications. Generates 10-100+ pages of realistic, cross-referenced, domain-appropriate synthetic data with proper formatting, metadata, and RAG-friendly structure. Keywords: synthetic data, dataset generation, dummy data, test data, mock documents, institutional knowledge, training data, sample documents, realistic data, batch generation, identity cards, resumes, invoices, bank statements, medical records, contracts, organizational documentation.
development
deep-thinking
Structured deep thinking for hard problems. Load when user says "deep think", "think deeply", "frame this", "what's the real question", "eigenquestion", "challenge this", "ultrathink", "stress test this", "apply a framework", "get expert feedback on this". Frames the problem (one-sentence problem, binding constraint, eigenquestion), then runs either stress-test (multi-perspective critique), mental framework (pre-mortem, decision-matrix, etc.), or expert review (persona-based feedback).
tools
generate-philosophy-doc
create philosophy doc, generate standards, build best practices guide.
tools
extract-requirements
Extract and organize requirements from meeting transcripts, documents, and other sources into structured domain files. Load when user says 'extract requirements', 'gather requirements', 'requirements from meetings', 'document requirements', or needs to consolidate requirements from multiple sources into organized documentation.
documentation
manager-review
Generate a structured manager review briefing for a team member. Load when user says 'manager review - [name]', 'prepare manager review', 'review prep for [name]', 'manager briefing', or '1:1 prep for [name]'. Pulls OKRs from Notion and cross-references with BEO data, Fathom meetings, and chat logs to produce an actionable, conversation-ready performance briefing with prioritized questions, risks, appreciation points, and suggested actions.
testing
interview-coach
Analyze interview transcripts to help managers improve their interviewing skills. Load when user says "review my interview with [name]", "assess my interview", "how's the candidate", "interview feedback", "interview coach", "hr round review with [name]", "technical round review with [name]", "final round review with [name]", or provides an interview transcript. Supports 3 interview types - HR (cultural fit), Technical (skill assessment), Final (career aspirations).
databases
prospect-brief
Research a target prospect and build a 6-slide commercial brief + cover email. Load when user wants to send something to a new account, mentions a company + contact person they want to reach out to, or asks to build a brief, outreach deck, or outreach package for a specific prospect. NOT for full CEO pitch decks (use ceo-pitch-creation) or detailed proposals (use proposal-creation).
development
operating-diagnostic
Build an Operating Diagnostic that maps a customer's back-office, ranks pain by ROI/complexity, shows where Beam agents sit, communicates how Beam delivers, and asks the team to validate the operation view BEFORE pricing is proposed. First artifact in the three-document sequencing (Diagnostic → Capability → Commercial) under Project 86's transformation thesis, also framed as the seed of an enterprise operating graph. Load when user mentions 'operating diagnostic', 'process diagnostic', 'transformation diagnostic', 'back-office map', 'Palantir-shape diagnostic', 'first-document gate before pricing', 'operating graph', 'enterprise ontology', 'discovery graph', 'seed graph', or asks for 'a diagnostic for [customer]'.
development
proposal-creation
Build a deep, Palantir-grade Phase-1 Engagement Proposal for an enterprise account: a ~34-page commercial + technical PDF (exec brief, operating map, line-by-line value derivation, architecture/security, delivery + team, production proof/case studies, and commercial terms with pricing at the end, plus SOW/SLA/security annexes). This is the Gate-3 priced artifact that follows the Operating Diagnostic. Config-driven: fill one account.py, tailor the bespoke analytical pages, render to PDF. Triggers on 'build a proposal for X', 'engagement proposal', 'commercial proposal', 'SOW proposal', '/proposal X'.
development
project-restructure
Advisory-first project structure pass. Recommends and optionally applies folder taxonomy improvements, subfolder grouping, and subproject boundaries after cleanup. Triggers: 'project restructure', 'restructure project', 'organize project structure', 'improve folder structure', 'structure pass'.
tools
qa-test-case-writer
Comprehensive UI test case generation for QA teams. Use when the user wants to write test cases for a feature, project, or webpage. Generates happy flow, edge cases, and security test cases for client-facing UI (not API tests). Creates test cases in Notion databases following standardized format with Test Case ID, Title, Pre-condition, Test Steps, Expected Result, Test Data, Post-condition, Priority (P0-P3), and status tracking. Triggers on requests like "write test cases for [feature]", "create QA tests", "generate test scenarios", or when user provides feature documentation to test.
tools
transcript-to-hubspot-sections
Extract Situation, Pain, Impact, and Integrations from sales transcripts. Load when user wants transcript structured for HubSpot or business sections.
data-ai
fathom-get-transcript
Fetch full transcript from Fathom using a meeting/recording ID. Load when user says "get transcript", "fetch transcript for [ID]", "fathom transcript", "download meeting transcript", or has a meeting ID and wants the full conversation text.
tools
google-slides
Create and manage Google Slides presentations. Load when user mentions 'google slides', 'slides', 'presentation', 'create presentation', 'add slide', or references creating/editing slide decks.
development
ultrathink
Critical analysis that stress-tests content from multiple perspectives before shipping. Load when user says "ultrathink", "ultrathink this", "stress test this", "challenge this content", "validate this", "critique this from different angles", or provides content (email, document, strategy, proposal, paper) asking for critical review. Examines content through recipient POV, critic lens, devil's advocate, first principles, and risk analysis, then synthesizes findings into actionable report.
development
beam-api-reader
Comprehensive Beam AI API reference with all endpoints, authentication, request/response schemas, usage examples, and full agent creation workflow. Load when user says "beam api", "api reference", "beam endpoints", "how to call beam api", "beam api docs", "create beam agent", "deploy beam agent", "build a beam agent", "create agent from yaml", or needs to interact with the Beam AI REST API or deploy a new agent.
development
hubspot-crm
All HubSpot CRM operations — contacts, companies, deals, engagements, search, associations (read and create default), and delete (archive). Load when user says 'list contacts', 'create deal', 'search companies', 'log call', 'log email', 'create note', 'create meeting', 'get associations', 'link contact to company', 'delete contact', 'update contact', 'update deal', 'hubspot search'.
testing
write-integration-ticket
Write API integration tickets in structured specification format. Load when user mentions 'write integration ticket', 'create integration spec', 'integration specification', 'API action ticket', or wants to document an API integration with endpoints, parameters, acceptance criteria, and error handling. Creates Linear tickets with Integration label assigned to Adeel Akram.
development
learn-beam-next
Deprecated compatibility alias. Learn advanced Beam/Prism concepts with the consolidated learn skill. Use `learn` with mode `advanced` instead of this legacy Beam Next learning skill.
tools
setup-memory
Deprecated compatibility alias. Deprecated alias for setup-goals; personalize Prism with role, goals, preferences, and useful work context. Use `setup` with mode `personalize` instead of this legacy Beam Next learning skill.
development
beam-get-task-details
Get detailed information about a specific Beam.ai task including status, parameters, node execution, and graph state. Load when user says "get task details", "task info", "show task", "task status", "what happened with task", or needs to inspect a specific task.
development
beam-feedback-automation
Universal feedback collection automation for ANY Beam agent. Load when user mentions 'feedback automation', 'create feedback sheet', 'agent feedback', 'evaluation sheet', 'feedback for [agent-name]', or provides a Beam agent ID/URL. Dynamically analyzes agents via API, generates smart feedback fields, creates Google Apps Script, deploys to Sheets, and manages complete feedback workflow.
tools
beam-retry-tasks
Retry failed Beam.ai tasks. Load when user says "retry tasks", "rerun failed tasks", "retry beam tasks", "resubmit tasks", or needs to re-execute failed agent tasks.
data-ai
beam-ai-slide-library
Use the Beam AI reusable slide library: individual HTML slide templates extracted from Beam Figma rebuilds, kept separate from deck themes and full deck templates. Load when the user asks for a slide library, specific Beam slide patterns, reusable Figma-inspired slides, Prism slide-library items, or slide-level HTML templates.
development
graph-slicer
Slice Beam.ai agent graphs into separate node files with topology-aware naming and Mermaid flowchart generation. Supports both exported JSON files with graphTools arrays AND direct Beam API responses with embedded toolConfiguration. Handles all fill types (static, linked, ai_fill) with appropriate rendering. Use when users want to split an agent graph into individual node files.
tools
beam-ai-deck-templates
Use Beam AI deck and report design packs, HTML templates, and curated examples to create sales decks, customer intro decks, RPO decks, and DIN A4 use-case proposal reports. Load when the user asks for Beam-branded presentation templates, Prism-compatible deck templates, Beam report templates, customer intro decks, commercial proposals, or reusable HTML deck/report examples.
development
allhands-prep
Prepare all-hands meetings for Beam. Load when user says 'prepare all-hands', 'all-hands prep', 'plan all-hands meeting', 'create all-hands agenda', or mentions preparing for company-wide meetings. Helps structure the meeting, draft talking points, create prep checklists, and follow Beam's all-hands guidelines.
testing
beam-figma-to-html-slides
Convert Beam Figma slide designs into high-fidelity, editable HTML presentation decks. Use when Codex is asked to audit Figma slides, extract slide templates, rebuild Beam slides as HTML decks, decide whether Figma imagery should be exported or rebuilt in HTML/CSS, create Beam/Prism-compatible deck templates, or improve fidelity of existing Beam HTML slide rebuilds.
development
create-roadmap
Create a prioritized roadmap of what to build next. Dynamic items based on your goals.
development
scope-of-work-creation
Create professional proposals and scopes of work (SOW) for client projects using phased approach with success gates. Load when user mentions "SOW", "scope of work", "proposal", "create proposal", "generate SOW", "write scope", "project proposal", "statement of work", or when preparing client proposals with phased deliverables, pricing, and success criteria. Structures proposals with clear phases, cost breakdowns, deliverables, success gates, out-of-scope items, and legal terms.
tools
project-cleanup
Project and docs cleanup with confidence-based decisions. Three modes: archive done docs, fix broken project references, and compress docs (merge related files, distill verbose content to rules-only, rename for clarity). Triggers: 'project cleanup', 'docs cleanup', 'clean up docs', 'archive done docs', 'fix project links', 'audit projects', 'compress docs', 'merge docs'.
testing
attio-connect
Manage contacts, notes, and lists in Attio CRM. Use when user says: 'add contact', 'update CRM', 'attio', 'create note', 'search contacts', 'add to list', 'look up in CRM', 'manage contacts', or references the Personal CRM project.
business
airtable-master
Shared resource library for Airtable integration skills. DO NOT load directly - provides common references (setup, API docs, error handling, field types) and scripts used by airtable-connect, airtable-query, and airtable-sync.
development
hubspot-connect
Connect and operate HubSpot CRM workflows for contacts, companies, deals, associations, emails, calls, notes, and meetings. Use when the user asks for HubSpot CRM setup or record operations.
testing
reduce-hallucination
Audit a Beam agent for hallucination risk using interrogation-science techniques — check prompts, input/output schemas, node boundaries, and (when available) real task evidence; then recommend prompt fixes, schema fixes, and graph-level additions (Code Executor validators, LLM cross-examiner nodes, abstention routing). STRICTLY READ-ONLY: never modifies, publishes, or deploys anything — output is an audit report with ranked recommendations. Trigger on "audit this agent for hallucination", "why is this agent extracting wrong data", "check this prompt for hallucination risk", "reduce hallucination", "审计幻觉", "为什么提取不准", "检查这个 agent 的幻觉风险", or when a user shows correction/complaint data from a production agent.
development
reduce-hallucination
Audit a Beam agent for hallucination risk using interrogation-science techniques — check prompts, input/output schemas, node boundaries, and (when available) real task evidence; then recommend prompt fixes, schema fixes, and graph-level additions (Code Executor validators, LLM cross-examiner nodes, abstention routing). STRICTLY READ-ONLY — never modifies, publishes, or deploys; output is an audit report with ranked recommendations. Load when user says "audit this agent for hallucination", "why is this agent extracting wrong data", "check this prompt for hallucination risk", "reduce hallucination", or shows correction/complaint data from a production agent.
development