
Response timeout pattern in real_estate (3 failures)
--- name: HCA Deal Intelligence engine: markdown version: 1 category: sales tags: [deals, M&A, scoring, investment] auto_trigger: "Score companies on sell-likelihood for middle-market M&A advisory" inputs: - name: entity_id type: string description: "Knowledge entity UUID to score" required: true --- ## Description Score companies 0-100 on sell-likelihood for middle-market M&A advisory. ## Scoring Rubric (0-100 total, weighted) | Category | Weight | Max Points | What to look fo
Response timeout pattern in accounting (16 failures)
Response timeout pattern in HR (16 failures)
Response timeout pattern in investment (16 failures)
Response timeout pattern in operations (14 failures)
--- name: Luna engine: agent platform_affinity: claude_code fallback_platform: gemini_cli category: personal_assistant tags: [whatsapp, copilot, business, email, calendar, knowledge, jira, github, competitor_intelligence] auto_trigger: "Personal assistant, business co-pilot, email, calendar, knowledge graph, competitor monitoring, task management" --- You are Luna — the user's senior chief of staff, business co-pilot, and the brain behind the AgentProvision platform. IMPORTANT: For the tenant_
--- name: AI Lead Scoring engine: markdown version: 1 category: sales tags: [leads, scoring, AI, qualification, BANT] auto_trigger: "Score leads for AI platform fit using the AI Lead Scoring rubric" inputs: - name: entity_id type: string description: "Knowledge entity UUID to score" required: true --- ## Description Score leads 0-100 based on likelihood of becoming a customer for an AI/agent orchestration platform. ## Scoring Rubric (0-100 total) | Category | Max Points | What
Google Workspace Admin SDK: Audit logs and usage reports.
Google Calendar: Manage calendars and events.
Google Chat: Manage Chat spaces and messages.
Google Docs: Append text to a document.
Google Drive: Manage files, folders, and shared drives.
Subscribe to Google Workspace events.
Gmail: Reply to a message (handles threading automatically).
Gmail: Show unread inbox summary (sender, subject, date).
Gmail: Watch for new emails and stream them as NDJSON.
Google Model Armor: Sanitize a user prompt through a Model Armor template.
Google People: Manage contacts and profiles.
Google Sheets: Read and write spreadsheets.
Google Tasks: Manage task lists and tasks.
Google Workflow: Announce a Drive file in a Chat space.
Manage customer support — track tickets, respond, escalate issues.
Export a Google Sheets spreadsheet as a CSV file for local backup or processing.
Retrieve and review responses from a Google Form.
--- name: HCA Deal Intelligence engine: markdown version: 1 category: sales tags: [deals, M&A, scoring, investment] auto_trigger: "Score companies on sell-likelihood for middle-market M&A advisory" inputs: - name: entity_id type: string description: "Knowledge entity UUID to score" required: true --- ## Description Score companies 0-100 on sell-likelihood for middle-market M&A advisory. ## Scoring Rubric (0-100 total, weighted) | Category | Weight | Max Points | What to look fo
--- name: Lead Scoring engine: tool tool_class: LeadScoringTool version: 1 category: sales tags: [leads, scoring, qualification, sales, BANT] auto_trigger: "Score or qualify a lead using configurable rubrics" inputs: - name: entity_id type: string description: "Knowledge entity UUID to score" required: true - name: rubric_id type: string description: "Scoring rubric ID (ai_lead, hca_deal, marketing_signal)" required: false --- ## Description Score entities 0-100 usin
--- name: Integral Business Support engine: agent category: support tags: [business, support, transactions, forex, troubleshooting, integral] version: 1 inputs: - name: message type: string description: User message or task required: true auto_trigger: "transaction, failed trade, delayed, FIX session, liquidity provider, trace, business impact, settlement, client, forex, trading issue" --- # Integral Business Support Agent — Operations Intelligence You are the Integral Business S
--- name: Aremko Receptionist slug: aremko_receptionist engine: markdown platform_affinity: gemini_cli fallback_platform: claude_code category: hospitality tags: [aremko, hospitality, bookings, reservations, spa, whatsapp] auto_trigger: "Reserva, disponibilidad, cabaña, tinaja, masaje, desayuno, alojamiento, hospedaje, Aremko, slots, horarios, precios" --- # Identidad — Recepcionista Aremko Eres la recepcionista virtual de **Aremko Spa & Cabañas** (Puerto Varas, Chile). Hablas con clientes que
--- name: Report Generation engine: tool tool_class: ReportGenerationTool version: 1 category: data tags: [reports, excel, charts, visualization, documents] auto_trigger: "Generate a report, Excel file, or document with charts and data" inputs: - name: report_type type: string description: "Type of report to generate" required: true - name: data type: string description: "JSON data for the report" required: true --- ## Description Generate structured reports with cha
Response timeout pattern in ecommerce (2 failures)
Response timeout pattern in real_estate (16 failures)
--- name: Lead Scoring engine: tool tool_class: LeadScoringTool version: 1 category: sales tags: [leads, scoring, qualification, sales, BANT] auto_trigger: "Score or qualify a lead using configurable rubrics" inputs: - name: entity_id type: string description: "Knowledge entity UUID to score" required: true - name: rubric_id type: string description: "Scoring rubric ID (ai_lead, hca_deal, marketing_signal)" required: false --- ## Description Score entities 0-100 usin
--- name: "Hello World" engine: "python" script_path: "script.py" inputs: - name: "user_name" type: "string" description: "The name of the person to greet." required: true --- ## Description A simple skill that returns a greeting. It's used as a proof-of-concept for the skills layer.
--- name: SQL Query engine: tool tool_class: SQLQueryTool version: 1 category: data tags: [sql, query, data, analysis, datasets] auto_trigger: "Run SQL query against connected datasets to retrieve and analyze data" inputs: - name: query type: string description: "SQL query to execute" required: true --- ## Description Execute SQL queries on connected datasets to retrieve and analyze data. Returns query results as structured data.
--- name: "Scrape Competitor SEO" engine: "python" script_path: "script.py" inputs: - name: "url" type: "string" description: "The full URL of the competitor's webpage to scrape (e.g., https://www.competitor.com)." required: true --- ## Description Accepts a URL and returns key on-page SEO information, including the page title, meta description, H1 tags, and all H2 tags. This is used to quickly analyze a competitor's content strategy and keyword focus.
--- name: "Hello World" engine: "python" script_path: "script.py" inputs: - name: "user_name" type: "string" description: "The name of the person to greet." required: true --- ## Description A simple skill that returns a greeting. It's used as a proof-of-concept for the skills layer.
--- name: Math & Calculations engine: tool tool_class: CalculatorTool version: 1 category: general tags: [math, calculator, compute, numbers] auto_trigger: "Calculate, compute, or do math operations" inputs: - name: expression type: string description: "Mathematical expression to evaluate" required: true --- ## Description Perform calculations and number crunching. Supports arithmetic, percentages, conversions, and complex mathematical expressions.
--- name: Entity Extraction engine: tool tool_class: EntityExtractionTool version: 1 category: automation tags: [extraction, entities, NER, text, knowledge] auto_trigger: "Extract people, companies, and concepts from text" inputs: - name: text type: string description: "Text to extract entities from" required: true --- ## Description Extract people, companies, locations, and concepts from text using NER. Identified entities can be stored in the knowledge graph for future reference
--- name: Math & Calculations engine: tool tool_class: CalculatorTool version: 1 category: general tags: [math, calculator, compute, numbers] auto_trigger: "Calculate, compute, or do math operations" inputs: - name: expression type: string description: "Mathematical expression to evaluate" required: true --- ## Description Perform calculations and number crunching. Supports arithmetic, percentages, conversions, and complex mathematical expressions.
--- name: AI Lead Scoring engine: markdown version: 1 category: sales tags: [leads, scoring, AI, qualification, BANT] auto_trigger: "Score leads for AI platform fit using the AI Lead Scoring rubric" inputs: - name: entity_id type: string description: "Knowledge entity UUID to score" required: true --- ## Description Score leads 0-100 based on likelihood of becoming a customer for an AI/agent orchestration platform. ## Scoring Rubric (0-100 total) | Category | Max Points | What
--- name: Luna engine: agent platform_affinity: claude_code fallback_platform: gemini_cli category: personal_assistant tags: [whatsapp, copilot, business, email, calendar, knowledge, jira, github, competitor_intelligence] auto_trigger: "Personal assistant, business co-pilot, email, calendar, knowledge graph, competitor monitoring, task management" --- You are Luna — the user's senior chief of staff, business co-pilot, and the brain behind the AgentProvision platform. IMPORTANT: For the tenant_
Response timeout pattern in research (16 failures)
Response timeout pattern in operations (14 failures)
Response timeout pattern in operations (16 failures)
Response timeout pattern in investment (16 failures)
Response timeout pattern in bookings (13 failures)
--- name: Aremko Receptionist slug: aremko_receptionist engine: markdown platform_affinity: gemini_cli fallback_platform: claude_code category: hospitality tags: [aremko, hospitality, bookings, reservations, spa, whatsapp] auto_trigger: "Reserva, disponibilidad, cabaña, tinaja, masaje, desayuno, alojamiento, hospedaje, Aremko, slots, horarios, precios" --- # Identidad — Recepcionista Aremko Eres la recepcionista virtual de **Aremko Spa & Cabañas** (Puerto Varas, Chile). Hablas con clientes que
--- name: Quick Statistics engine: tool tool_class: DataSummaryTool version: 1 category: data tags: [statistics, summary, data, overview] auto_trigger: "Generate statistics, summaries, or overviews of data" inputs: - name: dataset_id type: string description: "Dataset UUID to summarize" required: true --- ## Description Generate summaries and statistical overviews of datasets automatically. Provides key metrics, distributions, and data quality indicators.
--- name: Integral DevOps engine: agent category: devops tags: [devops, jenkins, nexus, ci-cd, pipeline, releases, integral] version: 1 inputs: - name: message type: string description: User message or task required: true auto_trigger: "build, deploy, release, pipeline, jenkins, nexus, artifact, promote, CI/CD, docker image" --- # Integral DevOps Agent — Release Operations You are the Integral DevOps agent, responsible for CI/CD pipeline management and release operations. ## You
Read data from two tabs in a Google Sheet to compare and identify differences.
Google Slides: Read and write presentations.
Handle HR workflows — onboarding, announcements, and employee comms.
--- name: Knowledge Search engine: tool tool_class: KnowledgeSearchTool version: 1 category: general tags: [knowledge, search, graph, entities, memory] auto_trigger: "Search the knowledge graph for people, companies, or concepts" inputs: - name: query type: string description: "Search query" required: true --- ## Description Search and browse the knowledge graph. Find people, companies, deals, and concepts previously extracted from conversations and documents.
Read and write Google Forms.
Create, organize, and distribute content across Workspace.
--- name: DevOps Agent engine: agent category: devops tags: [devops, ci-cd, pipeline, releases, jenkins, github-actions, nexus, artifactory, docker] version: 1 inputs: - name: message type: string description: User message or task required: true auto_trigger: "build, deploy, release, pipeline, jenkins, github actions, nexus, artifactory, artifact, promote, ci/cd, docker image, helm, kubernetes deploy" --- # DevOps Agent — Release Operations You are a DevOps engineer who runs CI/C
Gmail: Forward a message to new recipients.
Administer IT — monitor security and configure Workspace.
Organize research — manage references, notes, and collaboration.
Manage sales workflows — track deals, schedule calls, client comms.
Response timeout pattern in accounting (16 failures)
Response timeout pattern in accounting (16 failures)
Response timeout pattern in ecommerce (2 failures)
Response timeout pattern in operations (16 failures)
Response timeout pattern in real_estate (3 failures)
Response timeout pattern in real_estate (16 failures)
Response timeout pattern in research (16 failures)
Lead a team — run standups, coordinate tasks, and communicate.
--- name: Math & Calculations engine: tool tool_class: CalculatorTool version: 1 category: general tags: [math, calculator, compute, numbers] auto_trigger: "Calculate, compute, or do math operations" inputs: - name: expression type: string description: "Mathematical expression to evaluate" required: true --- ## Description Perform calculations and number crunching. Supports arithmetic, percentages, conversions, and complex mathematical expressions.
Google Workspace Events: Subscribe to Workspace events and stream them as NDJSON.
Google Calendar: Show upcoming events across all calendars.
Google Calendar: Create a new event.
Google Chat: Send a message to a space.
Google Classroom: Manage classes, rosters, and coursework.
Read and write Google Docs.
Google Drive: Upload a file with automatic metadata.
Google Workspace Events: Renew/reactivate Workspace Events subscriptions.
Gmail: Send, read, and manage email.
Gmail: Reply-all to a message (handles threading automatically).
Gmail: Send an email.
Manage Google Keep notes.
Manage Google Meet conferences.
Google Model Armor: Filter user-generated content for safety.
Google Model Armor: Create a new Model Armor template.
Google Model Armor: Sanitize a model response through a Model Armor template.
gws CLI: Shared patterns for authentication, global flags, and output formatting.
Google Sheets: Append a row to a spreadsheet.
Google Sheets: Read values from a spreadsheet.
Google Workflow: Cross-service productivity workflows.
Google Workflow: Convert a Gmail message into a Google Tasks entry.
Google Workflow: Prepare for your next meeting: agenda, attendees, and linked docs.
Google Workflow: Today's meetings + open tasks as a standup summary.
Google Workflow: Weekly summary: this week's meetings + unread email count.
Plan and manage events — scheduling, invitations, and logistics.
Manage an executive's schedule, inbox, and communications.
Coordinate projects — track tasks, schedule meetings, and share docs.
Add a list of attendees to an existing Google Calendar event and send notifications.
Create recurring focus time blocks on Google Calendar to protect deep work hours.
List and download all files from a Google Drive folder.
--- name: Entity Extraction engine: tool tool_class: EntityExtractionTool version: 1 category: automation tags: [extraction, entities, NER, text, knowledge] auto_trigger: "Extract people, companies, and concepts from text" inputs: - name: text type: string description: "Text to extract entities from" required: true --- ## Description Extract people, companies, locations, and concepts from text using NER. Identified entities can be stored in the knowledge graph for future reference
--- name: "Hello World" engine: "python" script_path: "script.py" inputs: - name: "user_name" type: "string" description: "The name of the person to greet." required: true --- ## Description A simple skill that returns a greeting. It's used as a proof-of-concept for the skills layer.
--- name: Marketing Signal Scoring engine: markdown version: 1 category: marketing tags: [marketing, engagement, intent, signals] auto_trigger: "Score leads based on marketing engagement and buying intent signals" inputs: - name: entity_id type: string description: "Knowledge entity UUID to score" required: true --- ## Description Score leads 0-100 based on marketing engagement, campaign response, and buying intent signals. ## Scoring Rubric (0-100 total) | Category | Max Point
--- name: Integral SRE engine: agent category: infrastructure tags: [sre, monitoring, alerts, infrastructure, integral] version: 1 inputs: - name: message type: string description: User message or task required: true auto_trigger: "infrastructure monitoring, alert investigation, server health, SSH operations, incident triage, runbook, haproxy, prometheus, database health, JBoss, messaging" --- # Integral SRE Agent — Technical Support You are the Integral SRE agent, a technical su
--- name: "Scrape Competitor SEO" engine: "python" script_path: "script.py" inputs: - name: "url" type: "string" description: "The full URL of the competitor's webpage to scrape (e.g., https://www.competitor.com)." required: true --- ## Description Accepts a URL and returns key on-page SEO information, including the page title, meta description, H1 tags, and all H2 tags. This is used to quickly analyze a competitor's content strategy and keyword focus.
--- name: SQL Query engine: tool tool_class: SQLQueryTool version: 1 category: data tags: [sql, query, data, analysis, datasets] auto_trigger: "Run SQL query against connected datasets to retrieve and analyze data" inputs: - name: query type: string description: "SQL query to execute" required: true --- ## Description Execute SQL queries on connected datasets to retrieve and analyze data. Returns query results as structured data.
--- auto_trigger: revenue,pending treatment, money on the table, or asks to audit the CDT codes in the CSV category: data chain_to: - "Accounting \u2014 Tool Handler (d62acf)" engine: python inputs: - description: he raw dental CSV export from the DSO practice management system name: csv_data required: false type: string name: dental_revenue_auditor script_path: script.py tags: - DSO - Revenue - CDT - Dental-Audit version: 1 --- ## Description dentifies high-value dental revenue opport
--- name: Report Generation engine: tool tool_class: ReportGenerationTool version: 1 category: data tags: [reports, excel, charts, visualization, documents] auto_trigger: "Generate a report, Excel file, or document with charts and data" inputs: - name: report_type type: string description: "Type of report to generate" required: true - name: data type: string description: "JSON data for the report" required: true --- ## Description Generate structured reports with cha
--- name: Innovus Terraform Infrastructure engine: agent platform_affinity: claude_code fallback_platform: codex category: infrastructure tags: [innovus, terraform, iac, aws, drift, modules, plan-review, policy-as-code] version: 1 tool_groups: [github, knowledge_readonly, drive, meta] inputs: - name: message type: string description: Terraform/IaC task, module review, drift question, plan review, or AWS infrastructure change required: true auto_trigger: "Terraform, tf plan, tfstate
--- name: Levi MDM PC9 Triage engine: agent platform_affinity: claude_code fallback_platform: codex category: infrastructure tags: [levis, mdm, pc9, s4, service-now, affiliate-activation, product-data] version: 1 tool_groups: [github, knowledge_readonly, drive, meta] inputs: - name: message type: string description: PC9, MDM, affiliate activation, S4, plant assignment, or ServiceNow evidence request required: true auto_trigger: "PC9, MDM, S4, Plant 2011, affiliate activation, drop
--- name: Innovus AWS Platform engine: agent platform_affinity: claude_code fallback_platform: codex category: infrastructure tags: [innovus, aws, devops-platform, platform-engineering, eks, ecs, rds, iam, observability] version: 1 tool_groups: [github, knowledge_readonly, drive, meta] inputs: - name: message type: string description: AWS platform task, onboarding question, incident, runbook, or architecture review required: true auto_trigger: "Innovus AWS, AWS platform, DevOps pla
--- name: Integral SRE Ops engine: agent platform_affinity: claude_code fallback_platform: codex category: infrastructure tags: [integral, sre, fxcw, jenkins, nexus, grafana, opentsdb, haproxy, alerts, rca] version: 1 tool_groups: [github, knowledge_readonly, drive, meta] inputs: - name: message type: string description: Integral SRE task, alert triage, RCA, Jenkins/Nexus/Grafana/OpenTSDB/HAProxy investigation required: true auto_trigger: "Integral, FXCW, OpenTSDB, Grafana, Jenkins
--- name: Business Support Agent engine: agent category: support tags: [business, support, transactions, troubleshooting, ops-intelligence] version: 1 inputs: - name: message type: string description: User message or task required: true auto_trigger: "transaction, failed transaction, delayed, business impact, settlement, customer issue, operational issue, ops support, trace order" --- # Business Support Agent — Operations Intelligence You are a business operations support special
--- name: Lead Scoring engine: tool tool_class: LeadScoringTool version: 1 category: sales tags: [leads, scoring, qualification, sales, BANT] auto_trigger: "Score or qualify a lead using configurable rubrics" inputs: - name: entity_id type: string description: "Knowledge entity UUID to score" required: true - name: rubric_id type: string description: "Scoring rubric ID (ai_lead, marketing_signal, or a tenant-defined rubric)" required: false --- ## Description Score e
--- name: SRE Agent engine: agent category: infrastructure tags: [sre, monitoring, alerts, infrastructure, on-call, observability, incident] version: 1 inputs: - name: message type: string description: User message or task required: true auto_trigger: "infrastructure monitoring, alert investigation, incident triage, on-call, server health, ssh, runbook, prometheus, grafana, opentsdb, latency, downtime, p1, p2, sev1" --- # SRE Agent — Incident Triage & Infrastructure Health You ar
--- name: Luna engine: agent platform_affinity: claude_code fallback_platform: gemini_cli category: personal_assistant tags: [whatsapp, copilot, business, email, calendar, knowledge, jira, github, competitor_intelligence] auto_trigger: "Personal assistant, business co-pilot, email, calendar, knowledge graph, competitor monitoring, task management" --- You are Luna — the user's senior chief of staff, business co-pilot, and the brain behind the AgentProvision platform. ## Anti-Hallucination Rule
--- auto_trigger: revenue,pending treatment, money on the table, or asks to audit the CDT codes in the CSV category: data chain_to: - "Accounting \u2014 Tool Handler (d62acf)" engine: python inputs: - description: he raw dental CSV export from the DSO practice management system name: csv_data required: false type: string name: dental_revenue_auditor script_path: script.py tags: - DSO - Revenue - CDT - Dental-Audit version: 1 --- ## Description dentifies high-value dental revenue opport
--- name: Levi SRE Platform engine: agent platform_affinity: claude_code fallback_platform: codex category: infrastructure tags: [levis, sre, ai-sre-platform, mdm, service-now, status, weekly-trackers] version: 1 tool_groups: [github, knowledge_readonly, drive, meta] inputs: - name: message type: string description: Levi's SRE platform task, weekly tracker, repo status, incident prep, or meeting briefing required: true auto_trigger: "Levi, Levi's, ai-sre-platform, SRE tracker, week
--- name: Knowledge Search engine: tool tool_class: KnowledgeSearchTool version: 1 category: general tags: [knowledge, search, graph, entities, memory] auto_trigger: "Search the knowledge graph for people, companies, or concepts" inputs: - name: query type: string description: "Search query" required: true --- ## Description Search and browse the knowledge graph. Find people, companies, deals, and concepts previously extracted from conversations and documents.
--- name: Knowledge Search engine: tool tool_class: KnowledgeSearchTool version: 1 category: general tags: [knowledge, search, graph, entities, memory] auto_trigger: "Search the knowledge graph for people, companies, or concepts" inputs: - name: query type: string description: "Search query" required: true --- ## Description Search and browse the knowledge graph. Find people, companies, deals, and concepts previously extracted from conversations and documents.
--- name: Marketing Signal Scoring engine: markdown version: 1 category: marketing tags: [marketing, engagement, intent, signals] auto_trigger: "Score leads based on marketing engagement and buying intent signals" inputs: - name: entity_id type: string description: "Knowledge entity UUID to score" required: true --- ## Description Score leads 0-100 based on marketing engagement, campaign response, and buying intent signals. ## Scoring Rubric (0-100 total) | Category | Max Point
--- name: Quick Statistics engine: tool tool_class: DataSummaryTool version: 1 category: data tags: [statistics, summary, data, overview] auto_trigger: "Generate statistics, summaries, or overviews of data" inputs: - name: dataset_id type: string description: "Dataset UUID to summarize" required: true --- ## Description Generate summaries and statistical overviews of datasets automatically. Provides key metrics, distributions, and data quality indicators.
Response timeout pattern in accounting (16 failures)
--- name: Luna Learn from Media engine: markdown category: meta tags: [learning, video, transcription, knowledge, meta] auto_trigger: "When the user sends a YouTube/Instagram/short-form video URL or asks you to 'learn this', 'study this clip', 'turn this into a skill', or otherwise convert media into an installable capability." inputs: - name: source_url type: string description: "URL of the media (YouTube, youtu.be, Instagram reel/post). Optional if attachment_path supplied." requ
--- name: Code Reviewer engine: agent platform_affinity: claude_code fallback_platform: codex category: engineering tags: [code-review, security, architecture, superpowers, regression, test-gaps] version: 1 tool_groups: [github, knowledge_readonly, meta] inputs: - name: message type: string description: A PR URL, PR number, branch name, or freeform review request required: true auto_trigger: "code review, PR review, diff review, regression check, security review, architecture revie
--- name: Substrate Sentinel engine: agent platform_affinity: claude_code fallback_platform: codex category: engineering tags: [security, tenant-isolation, audit, tool-scope, protocol-integrity, fail-closed] version: 1 tool_groups: [github, knowledge_readonly, meta] inputs: - name: message type: string description: A PR, plan, incident, or architecture question involving platform integrity required: true auto_trigger: "tenant isolation, tool scope, audit, fail closed, jwt, mcp auth
--- name: Luna engine: agent platform_affinity: claude_code fallback_platform: gemini_cli category: personal_assistant tags: [whatsapp, copilot, business, email, calendar, knowledge, jira, github, competitor_intelligence] auto_trigger: "Personal assistant, business co-pilot, email, calendar, knowledge graph, competitor monitoring, task management" tool_groups: [calendar, email, drive, data, reports, bookings, monitor, jira, github, workflows, skills, ecommerce, competitor, knowledge, meta, sales
Response timeout pattern in HR (16 failures)
Response timeout pattern in bookings (13 failures)