40 skills on the InTouch Hub
- API Health Monitor — Cluster failures across multiple HTTP endpoint health checks. AI groups failures by region, service, or status code and identifies likely root causes when failures share a pattern.
- AWS Cost Watcher — Daily AWS spend summary by service with anomaly detection. AI flags unusual spikes, identifies which services drove yesterday's cost, and projects month-end totals.
- Bid Estimator — Ballpark a bid for a defined scope of construction work using user-provided unit costs, crew rates, and overhead targets. Sanity check, not a substitute for a real estimate.
- Change Order Explainer — Plain-English breakdown of a change order — what scope changes, cost impact, time impact, who pays, and hidden clauses that often slip past PMs. For pre-signature review.
- Changelog Writer — Write a user-facing changelog entry from a list of merged commits/PRs.
- Commit Message Improver — Rewrite a vague commit message into Conventional Commits format using the diff for context. Picks type and scope from what the diff actually does, not just what the original message claimed.
- Comparable Property Analyzer — Compare a subject property against 3–8 comps. Summarize where the subject stands on price/sqft, days on market, condition adjustments. Flags outliers. Informational read for the listing agent or buyer's agent; not a substitute for a CMA or formal appraisal.
- Contract Clause Extractor — Extract key clauses (term, renewal, liability, termination) from contract text.
- CSI Spec Summary — Summarize a CSI MasterFormat spec section — title, scope, key requirements (Part 1 General, Part 2 Products, Part 3 Execution), submittal requirements, and common pitfalls. For PM/sub orientation; not a substitute for reading the actual section.
- CSV Anomaly Detector — Scan a CSV sample for anomalies (outliers, missing, type drift) and report findings.
- CSV Import Validator — Inspect a CSV file's headers and sample rows for schema problems, weird columns, missing required fields, encoding issues, and obvious data quality red flags before loading into a database.
- customer-followup — Process customer responses to follow-up messages after service/purchase. Route positive feedback to review requests, negative feedback to the owner for immediate action.
- daily-sales-reconciler — Process end-of-day sales photos from registers or POS systems. Extract totals, compare against expected ranges, flag discrepancies, and generate daily summaries for the owner.
- Error Postmortem Drafter — Generate a first-draft incident postmortem from a timeline + log dumps + optional chat thread. Produces Summary, Timeline, Root Cause, Resolution, Prevention, What-Went-Well, What-Didn't sections — blameless tone, ready for human edit.
- expense-tracker — Process expense receipts from any channel. Extract merchant, amount, date, category from receipt photos/PDFs. Enforce spending policies. Accumulate into expense reports. Submit for approval.
- GD&T Explainer (ASME Y14.5) — Explain a feature control frame in plain English — what it controls, the datum reference frame, the tolerance zone, how to measure it, and common gotchas. ASME Y14.5-2018 vocabulary; works from text input or a drawing image.
- GitHub Activity Digest — Weekly summary of commits, pull requests, and issues across watched GitHub repositories. AI groups by repo and theme, highlights stalled PRs and unresolved issues.
- Inbox Summary — Daily summary of unread email, categorized into Action Required, FYI, and Newsletter buckets. AI surfaces what actually needs a reply versus what can wait.
- Invoice Field Extractor — Extract structured fields (vendor, number, date, line items, total) from invoice text.
- invoice-processor — Extract structured data from invoice emails and PDF attachments. Returns vendor, amount, line items, due date, and payment terms. Flags unusual amounts or unknown vendors.
- Lead Classifier — Score and classify an inbound sales lead from its profile/message; suggest next action.
- Log Error Clusterer — Cluster a batch of error log lines into root-cause groups with counts and likely causes.
- Log Error Scanner — Cluster ERROR/FATAL log lines by signature, rank by severity and frequency, and surface the top issues from the last hour. AI groups stack-trace variants of the same root cause.
- Material Substitution Advisor — Suggest material substitutions when the specified material is unavailable or unsuitable. Returns 2-4 candidates with property comparisons, machining notes, cost ratios, lead times, and customer-approval implications (ECN/MRB/FAI requirements for aerospace/medical work).
- Meeting Notes Summarizer — Turn a raw meeting transcript into structured notes — action items with owners, decisions made, open questions, and key topics. Designed for the kinds of recordings Zoom, Teams, and OBS produce.
- Morning Briefing — Daily morning briefing combining today's weather, calendar events, and top news headlines into a single summary delivered at the time of your choosing.
- PDF Form Filler — Match structured data against a form template's labeled fields and fill the form. AI handles fuzzy field-name matching ("DOB" ↔ "dateOfBirth") and combines fields when needed (firstName + lastName → "Full Name").
- PR Diff Summary — Summarize a pull-request diff into a concise reviewer-facing changelog with risk flags.
- Quote Estimator (Machine Shop) — Produce a rough first-pass quote estimate (setup, cycle, material, OSP, suggested price, lead time, confidence) for a machined or fabricated part. Starting point for a senior estimator's review — not a final quote.
- Release Notes Drafter — Draft release notes grouped by Added/Changed/Fixed from a list of changes.
- Rental Yield Calculator — Compute cap rate, cash-on-cash return, NOI, DSCR, and ROI from user-provided purchase price, rent, financing, and expenses. Sanity check for an investment property; not investment advice.
- Repo Release Notes — Generate user-facing release notes from a git commit log between two tags. AI clusters commits into Features / Fixes / Improvements / Breaking Changes / Internal, drops chore commits, rewrites terse messages into prose.
- Sentiment Tagger — Tag a batch of messages/reviews with sentiment + theme and surface the notable ones.
- Social Post Drafter — Draft platform-appropriate social posts from a topic/brief with hashtags and variants.
- Specification Summary (AMS / ASTM / MIL-STD / etc.) — 60-second orientation on an industry spec — title, scope, key requirements, who invokes it, common pitfalls, verification trail, and related specs. For triage and starting research; not a substitute for the actual standard.
- SQL Query Explainer — Explain a SQL query in plain English and flag performance/correctness risks.
- Submittal Reviewer — Cross-check a submittal package (data sheet, product cut sheet, sample literature) against the spec excerpt it's submitted under. Flag missing data, spec mismatches, and non-equivalent product substitutions. Pre-screen for the PM before formal review.
- Support Email Triage — Triage an inbound support email: urgency, category, suggested owner, draft reply.
- Support Ticket Triager — Classify a support ticket by category, priority, and sentiment. Drafts a polite first-touch response (for human review) grounded in the ticket text and an optional knowledge base. Designed for the high-volume first-line support funnel.
- timesheet-collector — Collect employee work hours via any messaging channel. Extract hours per day, project/task notes, and overtime. Validate totals, flag anomalies, and generate weekly summaries for managers.