Coding & Dev Tools

natan-mohart Bundle

Computes probability-weighted outcomes across base, upside, and downside cases via a bundled calculator, and flags when the downside case breaches a stated minimum threshold like a cash floor or covenant level. Use whenever the user needs a scenario plan for a financial metric, wants to stress-test a plan against uncertainty, or has base/upside/downside cases with no probability weighting and no check against the actual minimum the business can survive.

natan-mohart Bundle

Tests projected financials against covenant thresholds quarter by quarter via a bundled calculator, computing headroom percentage and flagging the first quarter headroom drops below a warning threshold, not just the quarter of actual breach, so the lender conversation starts while options are still wide open. Use whenever the user needs to monitor covenant compliance, is projecting whether a covenant will hold under a forecast, or has covenant tracking that only reports pass or fail with no early-warning threshold before an actual breach.

natan-mohart Bundle

Computes the cash conversion cycle from DSO, DIO, and DPO via a bundled calculator, then ranks specific cash-release levers by actual dollar impact, turning "improve working capital" into a prioritized, sized list of concrete actions. Use whenever the user wants to free up cash trapped in the balance sheet, needs to compute or improve the cash conversion cycle, or has a working capital initiative with no ranked, dollar-sized list of what to actually do.

natan-mohart Bundle

Computes current, quick, cash, leverage, and interest-coverage ratios via a bundled calculator against stated healthy benchmark ranges, and prioritizes interest coverage and leverage flags over a low cash ratio alone, which is common and not automatically alarming. Use whenever the user wants a balance-sheet health read, needs liquidity and solvency ratios for a lender or board update, or has ratios computed with no benchmark comparison or sense of which flagged ratio actually matters most.

natan-mohart Bundle

Runs the close checklist against a target close date via a bundled calculator that tracks genuinely blocked tasks (not just late ones) and folds account reconciliation in as a close step, flagging any unreconciled variance over a stated materiality threshold. Use whenever the user is running or reviewing the month-end close, wants to speed up a close that consistently slips, or has reconciliation treated as a separate exercise disconnected from the close checklist it should be gating.

natan-mohart Bundle

Projects weekly cash balance from actual inflows and outflows via a bundled calculator, flags the exact week the balance would breach a stated minimum, and sizes the shortfall, instead of a monthly cash view that's too coarse to catch a short but real liquidity gap. Use whenever the user needs near-term cash visibility, wants a rolling 13-week cash view, or has only a monthly cash forecast that's too coarse to catch a specific week where cash gets genuinely tight.

natan-mohart Bundle

Models competitor and market response scenarios as a decision tree with real probabilities, computing expected value per strategic option through a bundled calculator that also flags when the highest-EV choice has a worst case the business might not survive. Use whenever the user wants to war-game a competitive move, needs a scenario-planning exercise, is choosing between options with genuinely uncertain competitor/market reactions, or has done "scenario planning" that was really just a written narrative with no probabilities or payoffs attached.

natan-mohart Skill

Writes a one-page decision memo (recommendation, the real alternatives considered, the specific ask, and the decision deadline) in the Amazon/Bezos narrative-memo tradition, engineered to get a genuine yes/no instead of vague "let's discuss" feedback. Use whenever the user needs a decision memo, needs to get a specific approval or resource ask decided by an executive or board, or has a document that describes a situation but never actually asks for anything specific.

natan-mohart Bundle

Sizes a market by triangulating independent top-down and bottom-up estimates (with a bundled calculator that flags disagreement), then maps demand against supply quality to find attractive, under-served white space. Use whenever the user wants a TAM/SAM/SOM, asks how big a market is, wants to evaluate a new market/segment/geography to enter, or is about to put a single market-size number in a deck without having cross-checked it.

natan-mohart Bundle

Scores strategic and execution risks on likelihood, impact, AND velocity (how fast a risk could hit) via a bundled calculator — catching fast-moving risks a standard 2-factor heat map underranks — then forces a named owner and mitigation for every high-priority item. Use whenever the user wants a risk register, risk assessment, or mitigation plan for a strategy or initiative, or has a risk list with no prioritization, owners, or mitigation actions attached.

natan-mohart Bundle

Prepares a negotiation by computing the actual Zone of Possible Agreement via a bundled BATNA calculator, and explicitly flags when no ZOPA exists so effort goes into restructuring the deal or walking away instead of grinding on a price gap that can't close. Use whenever the user is preparing for a negotiation (deal terms, salary, vendor contract, partnership), needs to set a walk-away price, or is about to negotiate without having named their BATNA or estimated the counterparty's.

natan-mohart Skill

Models rival capabilities, incentives, and constraints to predict their most likely next moves using game-theoretic response logic, then pre-commits your own response before they act. Use whenever the user asks what a competitor will do next, wants a competitive response plan, is about to launch something and wants to anticipate retaliation, or is comparing competitors on features alone instead of on capability and incentive.

natan-mohart Skill

Structures a strategy or recommendation into the Pyramid Principle (answer-first, MECE-supported) and a Situation-Complication-Resolution arc, so a senior audience gets the conclusion before the reasoning instead of a chronological walk through the analysis. Use whenever the user needs to turn analysis into an executive-ready narrative, deck, or presentation, or has a draft that walks through the work chronologically ("first we looked at X, then we looked at Y") instead of leading with the answer.

natan-mohart Skill

Generates a genuinely distinct set of strategic options (not variations on one idea), evaluates them on the same weighted criteria, and stress-tests the leading option against real-option/staged-commitment logic before recommending it. Use whenever the user wants to compare strategic paths, is choosing between build/buy/partner, needs a "should we do A or B" recommendation, or already has one preferred option and wants it validated rather than actually challenged.

natan-mohart Skill

Closes the loop that every other financial-case skill in this pack opens — checks realized results against the original business case, full-potential diagnostic, or synergy case line by line, and attributes the gap to assumption error, execution, or external shock. Use whenever the user wants a post-implementation review, wants to check if an initiative or deal actually delivered its promised value, or has a past business case, full-potential diagnostic, or synergy case that's never been checked against what actually happened.

natan-mohart Skill

Builds needs-based, mutually exclusive customer segments using Jobs-to-be-Done rather than demographic personas, then scores each segment on attractiveness and your actual right-to-win. Use whenever the user wants to segment customers, define an ICP, asks who to target, or has "personas" that describe demographics/psychographics but don't actually predict different buying behavior.

natan-mohart Bundle

Measures average span of control by management layer against benchmark via a bundled calculator, prices the annual cost of carrying more management layers than the organization needs, and flags narrow spans without assuming they're automatically wrong. Use whenever the user wants to review org structure for efficiency, is deciding whether to flatten layers or consolidate management roles, needs to size the cost of an over-layered organization, or has a headcount plan built without checking actual spans of control.

natan-mohart Bundle

Sizes the dollar value at stake between current performance and top-quartile benchmark on every relevant P&L and working-capital metric via a bundled calculator — the actual "full potential" framing used to size transformation programs, not a vague sense that "there's room to improve." Use whenever the user wants to size a transformation opportunity, benchmark performance against best practice, build the case for a performance improvement program, or has a transformation ambition with no quantified size of the prize behind it.

abchoudh-amd Skill

Stages changes and proposes atomic commits with short one-line subjects, executing only after explicit human approval. Use when the user asks Cursor to commit changes, create a commit, stage and commit work, write a commit message, generate a concise Git commit subject, or commit and push changes.

abchoudh-amd Bundle

Creates, resumes, edits, verifies, and marks ready exact ROCm pull requests with full Git OID checks, template and Systems PR Bot validation, draft-first creation, and canonical JSON receipts. Use when a user asks to create or update a rocprofiler-compute PR, publish one feature package PR, recover a partially created PR, inspect exact PR identity, or mark a verified draft ready.

abchoudh-amd Bundle

Smoke-tests rocprof-compute by running a curated YAML matrix of representative CLI flag combinations across the `profile` and `analyze` subcommands, passing on exit code within a per-case timeout. Use when a user asks to smoke-test rocprof-compute, sweep CLI flags, sanity-check a build's command surface, or verify a refactor did not break flag parsing or dispatch.

abchoudh-amd Bundle

Creates and publishes one approved executable rocprof-compute graph plan from a bounded aggregate Jira closure, through controlled questionnaire, formulation, approved low-level design, planner questions, and receipt-bound publication. Use for executable planning-only requests, literal `$plan-feature`, or exact active plan-feature state. Stops at a released `plan_ready` state and recommends `implement-feature`. Human-invoked only.

abchoudh-amd Bundle

Runs the rocprof-compute ctest suite against an existing build tree using a deterministic helper script. Use when a user asks to run tests, run ctest, verify a rocprof-compute build, or filter to a subset of tests.

abchoudh-amd Bundle

Builds and installs rocprof-compute through a CMake helper script with coverage and dev presets. Use when a user asks to build rocprof-compute, run coverage builds, or tune CMake build/install flags.

abchoudh-amd Bundle

Rewrites an existing document into plain language — shorter sentences, active voice, common words in place of jargon — and edits the file in place only after the human approves the specific changes. Use when a user asks to simplify, humanize, clarify, de-jargon, or plain-language a document, README, design doc, ticket description, or pasted prose, or to make writing easier to read.

abchoudh-amd Bundle

Maps a Python, C/C++/HIP, or CMake codebase into curated Mermaid diagrams saved as a code map under `./codemap/`. Use when a user wants to understand a codebase, map the code, see the call graph, draw a class diagram, show the build graph or CMake targets, or onboard onto an unfamiliar repo.

abchoudh-amd Bundle

Installs a ROCm nightly tarball from nightly.repo.amd.com and repoints the `current` symlink at it. Use when a user asks to install or update a ROCm nightly, list available builds, pin a version, or switch the active ROCm build.

appeeky Skill

When the user wants to edit an existing Appeeky-generated ad creative with natural language instructions. Use when the user mentions "edit ad creative", "change background", "bigger headline", "iterate on ad image", "tweak the ad", or mode edit for app ad creatives. For fresh angles see meta-ad-creative. For publishing see ad-creative-to-meta.

appeeky Skill

When the user wants multiple ad creative variants with different angles, styles, or image presets for A/B testing. Use when the user mentions "ad variants", "A/B creatives", "3 ad angles", "test different hooks", "creative batch for Meta", or "5 ad concepts". Uses Appeeky generate_app_ad_creative multiple times. For single creative see meta-ad-creative. For publishing see ad-creative-to-meta.

onfire-ai Skill

Resolve a company (from name, website, or LinkedIn URL) to a verified LinkedIn URL plus firmographics using the Onfire `match_company` tool. Use when you have any starting signal about a company — a name, a website, a CRM row, a LinkedIn URL you want to verify — and need a canonical identity. Also use as the upstream step before any `ai_prospecting` call that doesn't already have a company LinkedIn URL.

onfire-ai Skill

Detect month-by-month title movement (new hires and departures by job title) at any target company using `ONFIRE.EXPERIENCES_FULL` via `ask_onfire`. Use this skill whenever the user asks about: - "title movement" or "role movement" at a company - "new titles at [company]" or "what titles are they adding" - "titles that left [company]" or "what roles are disappearing" - "hiring signals for [company]" based on title patterns - "title changes at [company]" or "what roles are growing / shrinking" - detecting org changes, headcount shifts, or leadership gaps at a company - any question combining a company name with tracking who joined or left by title Trigger even if the user just pastes a LinkedIn company URL and asks "what's happening there?" or "show me movement." Works for any company — just needs a `company_linkedin_url` slug.

onfire-ai Skill

deanonymize_emails (atomic)

onfire-ai Skill

Office Segmentation

onfire-ai Skill

Search for companies directly from Onfire's LinkedIn company entity (the `company` entity, table ONFIRE.COMPANIES) using the `ask_onfire` tool. Use when the user wants company firmographics, technology adoption, funding data, or social URLs — phrases like "find fintech companies in Israel with 50-200 employees", "look up Northwind's company data", "which companies in our list are publicly traded", "get the GitHub URL for Artifex", or any company lookup that can be answered with structured filters on LinkedIn company data.

onfire-ai Skill

Find people who attended conferences, summits, webinars, or industry events using the `ask_onfire` tool against the `event_contact` / `event_company` entities. Use when the user wants to know who went to a specific event, find attendees from a target company, discover which events a person attended, or filter prospects by event participation — phrases like "who attended CloudCon 2025", "find people from Cloudforce at SecureCon", "which events did people from our target accounts go to", "show me Quadrant IT Symposium attendees", or any event attendance lookup.

onfire-ai Skill

community_messages_sentiment