206 Skills

When starting any non-trivial project, feature, or…

/apex-plan

Scopes a project or feature before any code is written. Reads the codebase and brief, challenges assumptions, and presents S, M, and L depth options with time and estimated token cost for each. You choose how deep to go before Apex dispatches any work to the team.

After completing a significant feature, before a p…

/apex-review

Reviews a completed body of work across all layers: API design, data model, security, observability, tests, and documentation. Catches the integration gaps that happen when specialists work in their own lane: not errors any one specialist would make, but the things that only appear when you look at the whole.

At the start of a work session to orient before di…

/apex-status

Reads git history, open PRs, recent commits, and codebase state to produce a CTO-level status report. Shows what is done, what is in progress, what is blocked, and what the highest-risk items are at this point in the project.

When inheriting a codebase from another team, when…

/apex-takeover

Takes ownership of an inherited or unfamiliar codebase by orchestrating parallel reconnaissance across all 14 specialists simultaneously. Produces a full system map, risk assessment, quick wins list, and recommended roadmap. This is the starting point when joining a project mid-stream.

When provisioning infrastructure for a new service…

/forge-infra

Builds production-grade infrastructure from scratch using Infrastructure as Code. Detects your cloud provider and target platform, then produces Terraform, Pulumi, or CDK output with compute, networking, storage, and IAM configured correctly from the start, not bolted on later.

When setting up network architecture for a new env…

/forge-network

Designs and builds networking infrastructure: VPCs with subnet strategy and CIDR planning, DNS configuration, load balancers with health checks and SSL termination, firewall rules with least-privilege ingress and egress. Documents every network decision with its rationale.

Before an enterprise sales deal that requires a se…

/forge-audit

Audits existing infrastructure for security misconfigurations, cost waste, and operational risks. Checks IAM permissions for over-privilege, public exposure on storage and databases, unencrypted resources, idle and oversized instances, and missing backup policies. Produces a prioritized finding list with remediation steps.

When cloud bills are rising without a clear explan…

/forge-cost

Analyzes cloud spend to identify what you are paying for and where you are overpaying. Finds idle resources, right-sizing opportunities for oversized instances, committed use discount gaps, and architectural changes that reduce cost without reducing capacity or reliability.

When a service is slow, timing out, or behaving un…

/forge-diagnose

Diagnoses runtime infrastructure problems by reading logs, metrics, and configuration: cold start latency, connection timeouts, autoscaling failures that leave services under-provisioned during spikes, network latency anomalies, and connection pool exhaustion. Identifies the actual cause rather than symptoms.

When taking over a project and need to understand …

/forge-recon

Infrastructure reconnaissance: inventories all cloud resources across accounts and regions, maps connections between services, identifies configuration drift between IaC definitions and what is actually running, and flags high-risk items. Produces a readable map of what you are running.

When a new service needs automated testing and dep…

/relay-pipeline

Builds a complete CI/CD pipeline from scratch. Detects your stack, test framework, and deployment target, then produces pipeline config for GitHub Actions, GitLab CI, Cloud Build, or CircleCI with proper stage sequencing, dependency caching, environment variable management, and deployment gates.

When containerizing a service for the first time. …

/relay-docker

Builds production-ready Dockerfiles with multi-stage builds to minimize image size, security hardening to drop root privileges and remove unnecessary packages, and Docker Compose for local development that matches production behavior. Includes .dockerignore and build cache optimization.

When deciding how to deploy a service safely witho…

/relay-deploy

Sets up a deployment strategy with explicit rollback procedures: rolling updates for zero downtime, canary releases for gradual traffic shifting with automatic rollback on error rate threshold, or blue-green for instant cutover. Includes smoke test definitions and deployment gates.

When CI is slow enough that developers are context…

/relay-audit

Audits an existing CI/CD pipeline for performance bottlenecks that slow developer feedback loops, security gaps like hardcoded secrets or over-permissioned service accounts, reliability issues like missing retries, and developer experience problems like cryptic failure messages.

When taking over a project and need to understand …

/relay-recon

Maps the full deployment chain from commit to production: all triggers, build steps, test execution stages, artifact publishing, environment promotion, and deployment steps, with risk assessment for each stage. Documents what happens when each step fails.

When building a new REST or GraphQL API from scrat…

/spine-api

Designs and builds an API contract-first: the interface is defined and agreed before implementation begins. Produces endpoint specs with request and response schemas, validation rules, auth requirements, error codes, and pagination patterns, then implements the full API against that spec.

When starting a new backend service or microservic…

/spine-service

Bootstraps a new backend service with everything production requires from day one: config management via environment variables with validation at startup, health check and readiness endpoints, graceful shutdown that drains in-flight requests, structured logging, and service discovery registration.

SpinePlanning

/spine-design

Produces a system design document before implementation begins: component breakdown, data flow diagrams, API contracts between services, failure modes with fallback behavior, and scaling strategy. Written to be reviewed and challenged before any code is written.

When architecting a new system or feature that inv…

When specific endpoints are slow under normal load…

/spine-perf

Finds and fixes performance bottlenecks in existing backend services. Diagnoses N+1 query patterns, missing database indexes, synchronous I/O that should be async, cache layer gaps, and connection pool exhaustion. Produces a prioritized list of fixes with the expected impact of each.

Before launching a new API to external users or th…

/spine-review

Reviews a backend codebase or API against production readiness criteria: REST convention adherence, authentication and authorization coverage, input validation completeness, error handling for all failure modes, pagination on every list endpoint, rate limiting, and test coverage of the critical path.

When taking over a backend codebase that has no do…

/spine-recon

Backend reconnaissance: maps all API routes, middleware chain, authentication setup, data model, external service dependencies, and codebase structure. Assesses code quality, test coverage, and identifies the highest-risk areas. Produces a readable inventory before making changes.

When designing a new data model from scratch and w…

/flux-schema

Designs a database schema with proper normalization, appropriate denormalization for high-frequency read paths, indexes for the expected query patterns, constraints that enforce data integrity at the database level, and migration files ready to run. Explains the reasoning behind every design decision.

Whenever making schema changes on a live productio…

/flux-migrate

Writes zero-downtime database migrations: reversible steps designed to run while the application is serving traffic, safe for concurrent reads and writes during the migration window, with an explicit rollback plan for every step. Handles column renames, backfills, and index creation on large tables safely.

When moving data between systems on a schedule. Wh…

/flux-pipeline

Builds a data pipeline from source to destination: extraction from APIs, databases, or files; transformation with type coercion, deduplication, and business logic; loading into the target store; with error handling, retry logic, dead letter queues, and scheduled or event-driven execution.

When specific queries are slow in production or st…

/flux-query

Optimizes slow database queries by reading execution plans, identifying missing indexes, rewriting inefficient joins and subqueries, and suggesting caching strategies where appropriate. Explains why each change helps so you can apply the same reasoning to future query problems.

When dashboards are showing unexpected numbers and…

/flux-health

Audits data quality and pipeline health across multiple dimensions: freshness of data in each table versus expected SLA, schema drift between environments, null rates on fields that should never be null, orphaned records from missing foreign key constraints, and execution status of all scheduled pipelines.

When taking over a project with a complex or undoc…

/flux-recon

Database reconnaissance: full inventory of schema structure and table relationships, migration history, data volume per table, backup configuration and last successful backup, connection pooling setup, and observed query patterns from slow query logs. Identifies what databases the system uses and where the risks are.

Before a production launch with external users. Be…

/warden-audit

Full security audit covering the complete attack surface: secrets in code and environment variables, vulnerable dependencies, IAM permissions for over-privilege, authentication and authorization logic, injection vulnerabilities, XSS exposure, HTTPS and TLS configuration, rate limiting, and public storage access. Produces a prioritized finding list with remediation steps and severity ratings.

When a service is being exposed to the public inte…

/warden-harden

Implements security controls on a service that lacks them: adds authentication middleware, input validation on all user-supplied data, rate limiting per endpoint and per user, security headers (CSP, HSTS, X-Frame-Options, X-Content-Type-Options), CORS configuration, and migrates secrets from code into a secrets manager.

When setting up cloud permissions for a new system…

/warden-iam

Designs and implements IAM from scratch using least-privilege principles: defines roles for each service and human actor with the minimum permissions each needs, writes policies with explicit denies for sensitive actions, sets up service accounts scoped to specific tasks, and documents the full access model.

When taking over a system and need to understand i…

/warden-recon

Security reconnaissance: inventories secrets management approach, IAM policies and their effective permissions, authentication methods per service, encryption at rest and in transit status, dependency vulnerability scan results, audit logging coverage, and known compliance gaps. Produces a security posture report with risk ranking.

When a service has no observability and you cannot…

/vigil-instrument

Instruments a service with production-grade observability: structured logging with consistent JSON fields and trace context, RED metrics (request rate, error rate, duration histograms) exported to Prometheus or your metrics stack, distributed tracing spans via OpenTelemetry, and health check and readiness endpoints.

When setting up alerting on a new service and want…

/vigil-alert

Builds alerting rules grounded in SLOs with paired runbooks for every alert. Defines error budgets, creates alert conditions at the right thresholds to minimize false positives, and writes runbooks with investigation steps and specific remediation options so the on-call engineer knows exactly what to do when paged.

When something is broken in production and you nee…

/vigil-incident

Leads production incident response: reads logs, metrics, traces, and recent deploy history to diagnose the problem systematically, identifies root cause rather than just symptoms, proposes a fix with rollback options, and documents findings in a structured post-mortem format.

Before a launch to verify the service is observabl…

/vigil-check

Audits monitoring coverage against actual service behavior: checks whether every critical user path has a corresponding alert, identifies monitoring gaps in background jobs, async processors, and queues, and produces a gap-prioritized action plan for improving coverage before problems occur.

When taking over a service and need to understand …

/vigil-recon

Observability reconnaissance: inventories all monitoring, alerting, logging, and tracing infrastructure. Maps what is covered, what is completely blind, and what alerts exist with their current thresholds. Produces a monitoring coverage report that identifies the highest-risk gaps sorted by impact.

When building a new page or web application from s…

/prism-ui

Builds a production-ready UI from scratch: pages with real data fetching, skeleton loading states, error boundaries, empty states, responsive layout across breakpoints, and keyboard navigation. Follows accessibility standards and uses the project's existing component patterns rather than introducing new ones.

When adding a new reusable component to a design s…

/prism-component

Builds a reusable, accessible, typed UI component with every state handled: default, hover, focus, active, disabled, loading, error, and any domain-specific states. Includes prop documentation, usage examples, keyboard interaction patterns, and ARIA attribute coverage.

When engineers or operations teams need a UI for d…

/prism-dashboard

Builds an internal dashboard with data tables that support sorting, filtering, and pagination; detail drawer or modal views; CRUD forms with validation; and search. Designed for internal tooling use cases where density and function matter more than marketing aesthetics.

When frontend performance is visibly slow or Light…

/prism-audit

Frontend audit: analyzes bundle size to identify what is driving it, checks for unused dependencies that should be removed, runs an accessibility audit against WCAG standards, measures Lighthouse performance metrics, and reviews component quality for consistency with the design system.

When taking over a frontend codebase that is unfam…

/prism-recon

Frontend reconnaissance: maps the component tree, routing structure, state management approach, build configuration, and third-party dependencies. Assesses code quality, component consistency, and technical debt. Produces a readable picture of the frontend before making changes that could have wide impact.

When building a prediction, classification, or reg…

/cortex-model

Builds an end-to-end ML pipeline: data ingestion with validation, feature engineering with a feature store, model training with cross-validation and hyperparameter tuning, evaluation against a held-out test set, and deployment to a serving endpoint with monitoring. Chooses the right model type for the problem.

When building AI features where prompt quality dir…

/cortex-prompt

Designs, versions, and evaluates prompts for LLM-powered features. Builds evaluation suites that test prompts against a representative set of inputs so quality regressions are caught before they reach users. Applies systematic prompt engineering techniques and documents what each version changes and why.

When adding an AI feature to an existing product f…

/cortex-integrate

Integrates an LLM into a production service: API client with retry and timeout handling, streaming response support, semantic caching to avoid redundant API calls, fallback provider configuration, and cost controls including token budget enforcement and request rate limiting per user.

When model outputs seem worse after a retraining r…

/cortex-eval

Evaluates a deployed model or LLM integration for performance issues: checks for accuracy degradation against a reference dataset, data distribution drift that may explain behavior changes, latency regression compared to baseline, and cost increases from token usage changes. Produces a health report with recommended actions.

When taking over a project that has existing ML in…

/cortex-recon

ML and AI reconnaissance: inventories all models in production, LLM integrations and their API usage, training pipelines and their schedules, data sources and their freshness, evaluation setups, and monitoring coverage. Identifies technical debt, models running without evaluation, and unmonitored AI features.

When starting a new mobile application and want a …

/touch-app

Scaffolds a production-ready mobile app from day one: navigation structure (stack, tab, drawer), API client with authentication and token refresh, push notification registration, offline storage setup, error handling patterns, and CI pipeline. Available for React Native, Flutter, and native iOS and Android.

When adding a new screen or user-facing capability…

/touch-feature

Builds a complete mobile feature: screen layout and navigation integration, state management and data fetching, API integration with loading and error states, offline handling with local cache, deep link support, and push notification handling. Follows the app's existing patterns.

When a mobile app needs automated release infrastr…

/touch-release

Sets up the full mobile release pipeline: Fastlane lanes for building, code signing, and App Store and Play Store submission; CI integration for automated builds on merge; beta distribution via TestFlight or Firebase App Distribution; versioning strategy; and store listing setup.

Before an App Store or Play Store submission to ca…

/touch-audit

Mobile app audit: measures app startup time and identifies what is causing it, analyzes app binary size, checks crash reporting coverage and recent crash rates, reviews App Store and Play Store compliance requirements, runs an accessibility check, and assesses offline behavior edge cases.

When taking over a mobile codebase that is unfamil…

/touch-recon

Mobile reconnaissance: maps the app's tech stack, navigation architecture, state management approach, API integration layer, third-party SDK inventory and their purposes, build configuration, code signing setup, and overall code quality. Produces a health assessment with risks flagged.

When starting a new embedded or IoT device project…

/volt-firmware

Scaffolds a complete embedded firmware project for a target MCU: RTOS task structure with stack size planning, peripheral initialization sequences, power management state machine, communication stack (BLE, MQTT, or serial protocol), OTA update client, and CI build configuration. Available for ESP-IDF, STM32 HAL, Zephyr, and Arduino frameworks.

When integrating a new sensor, peripheral, or comm…

/volt-driver

Writes a device driver or protocol handler with a clean HAL abstraction layer: I2C and SPI sensor drivers with interrupt-driven I/O, BLE GATT service implementations, MQTT client with reconnection logic, or Modbus RTU handlers. Designed to be unit-testable and portable across hardware revisions.

When devices are deployed in the field and need re…

/volt-ota

Builds a complete OTA firmware update system: dual-partition A/B storage with partition manager configuration, firmware image signature verification to prevent unauthorized updates, download with resume capability on connection loss, atomic partition swap, and automatic rollback if the new firmware fails to boot successfully.

When taking over a firmware project without docume…

/volt-recon

Firmware reconnaissance: inventories the MCU and full peripheral set, maps RTOS tasks with their stack usage and priorities, documents communication protocols and data flows, assesses OTA system readiness, audits power management implementation, and produces code quality risk flags for inherited firmware.

When you need a visual architecture overview for a…

/atlas-map

Reads the codebase to produce C4-style architecture diagrams in Mermaid: Level 1 context diagram showing system boundaries, external actors, and data flows; Level 2 container diagram showing services, databases, and how they connect. Adds written descriptions of each component's responsibility.

After any significant technical decision: choosing…

/atlas-adr

Writes an Architecture Decision Record for a significant technical choice: states what was decided, what alternatives were seriously considered, what the tradeoffs are between options, and what the consequences of this decision are for future work. Follows a consistent template that future engineers can read and understand.

When onboarding a new team member and you do not w…

/atlas-onboard

Generates comprehensive onboarding documentation written for a day-one engineer who knows nothing about the project: what the system does and why it exists, how to run it locally step by step, where things live in the codebase, key architectural decisions and their rationale, how to deploy safely, and who to ask when stuck.

When a Vigil, Forge, Apex, or other agent output e…

/atlas-report

Renders the output from any Tonone agent as a styled HTML page and opens it in the browser. Handles all the layout, sections, tables, and Mermaid diagram rendering. Used when findings are too detailed for the terminal: takeover reports, audit results, architecture maps, and any analysis that benefits from formatted presentation.

After completing a significant feature, sprint, or…

/atlas-changelog

Maintains structured CHANGELOG files after agent work completes. Appends entries in standard Keep a Changelog format (Added, Changed, Fixed, Removed, Security) with enough context to understand what changed and why, so a reader can understand the change without reading the git diff.

For major releases or product launches where stake…

/atlas-present

Produces a polished HTML presentation page and Obsidian Canvas for major project milestones: product launches, post-takeover system briefings, migration completion summaries, and release overviews targeted at non-technical stakeholders. The output is a shareable URL, not a terminal dump.

When taking over a project and need to understand …

/atlas-recon

Documentation reconnaissance: finds all documentation in the repository and external links, assesses accuracy and freshness against the current codebase state, maps coverage gaps by area, and identifies critical knowledge that exists only in someone's head and nowhere in writing.

When building a new analytics dashboard for a prod…

/lens-dashboard

Builds an analytical dashboard where every chart answers exactly one question for one specific audience. Detects your data stack (Metabase, Grafana, Streamlit, or Chart.js), designs a metric hierarchy, and implements the dashboard with proper labeling, filtering, drill-down capability, and data refresh configuration.

When starting a new product or feature and need to…

/lens-metrics

Defines a metrics framework from the top down: North Star metric with rationale for why it is the right measure of success, supporting input metrics with precise definitions, SQL queries for each metric, ownership assignment to a specific person or team, review cadence, and action triggers specifying what to do when each metric moves.

When stakeholders need recurring metric reports de…

/lens-report

Builds an automated reporting pipeline: SQL queries for each metric, scheduling logic (daily, weekly, monthly), delivery via Slack or email with proper formatting, threshold alerts for unusual values that need attention, and historical comparison tables. Low-maintenance once set up.

When analytics has grown organically and nobody is…

/lens-audit

Reviews existing analytics infrastructure: finds all dashboards and reports, checks usage data to identify which ones are actually opened versus ignored, verifies that metric definitions are documented and consistent across tools, and recommends what to keep, what to consolidate, and what to deprecate.

When taking over a product and need to understand …

/lens-recon

Analytics reconnaissance: inventories all analytics tools and integrations, maps which events and metrics are currently tracked and where they are defined, assesses data freshness against business requirements, documents metric definitions where they exist, and produces a coverage map showing what is visible and what is a blind spot.

When starting a new project and want to establish …

/proof-strategy

Designs a test strategy for a project: analyzes what tests already exist and at what layer, identifies critical user paths with no coverage, recommends the right mix of unit, integration, and E2E tests given the project's specific risk profile and release cadence, and produces a testing plan with priority order and rationale.

When critical user flows like sign-up, checkout, o…

/proof-e2e

Builds E2E test suites with Playwright, Cypress, or Selenium focused on user journeys rather than implementation. Covers the critical paths that must work for the product to deliver its core value, with stable selectors that survive UI changes, proper async waits, and CI integration that runs on every merge.

When an API needs systematic test coverage before …

/proof-api

Builds API test suites covering endpoint behavior (correct status codes, response schema validation, error cases), contract testing to catch breaking changes before they reach consumers, and load testing to establish baseline performance at expected traffic levels. Works for REST, GraphQL, and gRPC.

When CI is unreliable enough that team members hav…

/proof-audit

Audits test suite health: identifies flaky tests and their root causes, finds tests that take disproportionately long and explains why, measures coverage gaps on critical code paths, and reviews testing patterns for anti-patterns like testing implementation details or using inadequate assertions that pass on wrong behavior.

When taking over a project and need to understand …

/proof-recon

Testing reconnaissance: inventories all test files, testing frameworks, and configuration; measures coverage per module with a critical path focus; checks CI integration and test execution time; and assesses overall testing maturity. Produces a map of what is tested, how thoroughly, and what the biggest coverage gaps are.

When the team creates new services frequently and …

/pave-golden

Builds a golden path template for a specific service type: includes project scaffolding, local dev setup with one-command startup, CI pipeline, deployment configuration, and monitoring instrumentation baked in from the start. New services created from it start correctly and consistently.

When local setup is undocumented, takes hours, or …

/pave-env

Sets up local development environments with devcontainer configuration, Docker Compose service definitions that match production, and one-command startup so every developer has a consistent environment from day one. Addresses the 'works on my machine' problem systematically rather than with workarounds.

When the team does not have a clear picture of wha…

/pave-catalog

Builds a service catalog: inventories all services with their owners, descriptions, tech stacks, upstream and downstream dependencies, health status endpoints, and documentation links. Can be implemented in Backstage, a static site, or a structured document depending on team size and tooling.

When developers are slowed down by tooling but it …

/pave-audit

Audits developer experience by measuring the things that slow developers down: time to first working local environment, build times at different cache states, deployment friction from commit to production, test execution speed, and time to first production deploy for a new feature. Uses DORA metrics as a framework.

When taking over responsibility for developer expe…

/pave-recon

Platform engineering reconnaissance: inventories all developer tooling, build systems, environment configurations, local development setups, internal CLIs and scripts, and deployment workflows. Produces a picture of how developers currently work and where the friction points and inconsistencies are.

When planning a quarter and the backlog needs to b…

/helm-plan

Prioritizes a product backlog using scoring frameworks: RICE (Reach x Impact x Confidence divided by Effort), ICE scoring, or Kano model categorization. Produces a ranked backlog with written rationale for every item's position, an explicit list of items being deferred this cycle, and confidence levels on each estimate.

When a feature idea needs to become a spec before …

/helm-brief

Turns a feature idea or user problem into a structured product brief: problem statement with supporting evidence, target user definition (not just demographic, but specific context), proposed solution with alternatives considered, success criteria as measurable outcomes rather than activities, and explicit out-of-scope items. The output is a document Apex can act on without a follow-up meeting.

When a product brief is complete and ready for eng…

/helm-handoff

Validates a finalized product brief for completeness and handoff-readiness, then packages it for Apex to scope and execute. Checks that success criteria are measurable, out-of-scope items are explicit and agreed, all necessary context is present, and the brief is clear enough that nothing will be lost in translation during engineering scoping.

When you have completed user interviews and need t…

/echo-interview

Synthesizes user interview notes, customer feedback documents, or support ticket samples into structured insights. Extracts Jobs-to-Be-Done statements, identifies recurring pain points with frequency and severity, and produces a structured insight report with supporting evidence quotes. Works from any raw input you provide.

When activation is below target and it is unclear …

/lumen-funnel

Analyzes a product funnel to find where users drop off and why. Maps conversion rates at each step, compares cohorts by acquisition channel and signup date, identifies the highest-impact drop-off points, and produces specific recommendations for activation, onboarding, and conversion improvements.

When a feature needs a flow diagram before visual …

/draft-flow

Maps a user flow as a Mermaid diagram: the sequence of screens and decision points a user moves through to accomplish a specific goal, with error states, branch conditions, and explicit entry and exit points. Grounded in the job the user is trying to accomplish rather than the features being built.

When launching a new product that needs a visual i…

/form-brand

Creates brand identity foundations: color palette with primary, secondary, and semantic tokens with usage rules; type system with scale, pairing rationale, and line height specifications; iconography style direction; and style guide documentation. Every choice is tied to a brand principle so the reasoning can be applied consistently to new surfaces.

When planning a quarter or half-year and need prio…

/crest-roadmap

Builds a strategic product roadmap with RICE-scored items, an explicit now/next/later structure, competitive context explaining why certain bets are being made over others, OKR alignment for each initiative, and a documented not-now list with reasoning. Designed to survive tough questions from engineers, stakeholders, or investors.

When launching a product and the team cannot agree…

/pitch-position

Writes a positioning statement following the established template: target customer, problem category, key benefit, and differentiation from alternatives. Produces a brief that can serve as the strategic anchor for all marketing copy, sales messaging, and product decisions about who the product is for.

When new user activation rates are below benchmark…

/surge-activation

Maps the activation funnel from first use to the moment the product delivers its core value (the aha moment). Identifies each step in the journey, measures drop-off rates at each step, and produces an ordered list of activation improvements ranked by expected impact and implementation effort.

At the start of a session on a project you have no…

/apex-recon

Engineering lead reconnaissance: reads the codebase structure, git history, open work, and dependency graph to produce a focused inventory before planning begins. Faster and lighter than a full takeover, designed for orientation rather than comprehensive handoff documentation.

When asked 'how do I use tonone', 'what can tonone…

/tonone-onboard

Guided onboarding tour of tonone's 23 agents, key skills, and worktree sessions. Two paths: expert (~90 sec) and newcomer (~8 min). Cross-agent — not tied to any single specialist.

Before designing any security-sensitive feature: a…

/warden-threat

Threat modeling using the STRIDE framework: identifies attack surfaces by mapping data flows, defines trust boundaries between components, enumerates threat actors and their capabilities, and lists mitigations for each identified threat. Produces a threat model document with residual risk assessment.

When a battery-powered device is not meeting its b…

/volt-power

Analyzes the firmware power budget: maps all sleep modes and their wake sources, audits power state machine transitions for missing low-power states, measures radio duty cycles for BLE and WiFi including advertising intervals and connection parameters, and estimates battery life under different usage pattern scenarios.

When product wants to include something that engin…

/helm-arbiter

Runs a structured scope arbitration session when product and engineering cannot agree on what is in or out of scope for a feature or sprint. Hears both positions, applies tradeoff analysis with explicit criteria, produces a documented decision with rationale, and specifies an escalation path if the decision is later contested.

Before starting a new product initiative on a prod…

/helm-recon

Reads existing product briefs, research documents, strategy docs, roadmap files, and prior agent outputs to build a complete picture of the product before writing anything new. Identifies what has already been decided, what work is currently in progress, and where the gaps are in current product thinking.

When entering a market and need to understand the …

/crest-compete

Produces a competitive analysis: feature parity grid against named competitors showing where you are ahead, behind, or missing entirely; a positioning 2x2 on the two axes most relevant to the buying decision; and a classification of features as table stakes (required to compete) versus differentiators (reasons to choose you). Grounded in publicly available information.

At the start of a planning cycle when the team nee…

/crest-narrative

Writes a strategic narrative document: frames the product direction as a set of explicit bets, explains the reasoning and evidence behind each bet, makes tradeoffs visible rather than hiding them, and writes it at a level of clarity that an engineer or investor can understand without product context.

When setting goals for a quarter or year and want …

/crest-okr

Designs an OKR framework: Objectives framed as outcomes rather than activities, Key Results that are measurable, time-bound, and genuinely ambitious, an input metrics tree showing how each KR connects to the North Star, and a review cadence. Avoids the failure mode of KRs that describe work to be done rather than change to be achieved.

Before starting any roadmap, competitive analysis,…

/crest-recon

Reads existing strategy documents, roadmaps, OKRs, competitive materials, and prior product team outputs to establish the current strategic context before any planning work begins. Produces a summary of what has been decided, what is still in progress, and what strategic questions remain open.

When redesigning navigation and need a structural …

/draft-ia

Designs information architecture: navigation structure, content hierarchy, sitemap, and taxonomy for a product or feature set. Maps where every page and section lives and how users navigate between them. Produces a Mermaid sitemap and written rationale for every structural decision.

When a feature needs a lo-fi layout specification …

/draft-wireframe

Produces screen-level wireframes: text-based layouts and Mermaid diagrams showing content hierarchy, component placement, primary and secondary actions, and interaction annotations. Specifies structure and behavior without visual design, so engineering can estimate and Form can design with a shared structural reference.

When users are dropping off at a specific point in…

/draft-review

Evaluates an existing flow or UI against established usability heuristics: Nielsen's ten heuristics, Fitts's law for touch targets, cognitive load principles. Identifies specific friction points with supporting rationale, rates their severity, and recommends concrete, implementable fixes. Not aesthetic opinion, structured usability analysis.

Before designing a new flow on a product you have …

/draft-recon

Scans the frontend codebase to map existing routes, page components, navigation structure, and interaction patterns. Produces a UX map of what currently exists, including inconsistencies and patterns that have drifted from the intended design, before designing anything new.

When a large volume of qualitative feedback needs …

/echo-feedback

Clusters and synthesizes qualitative feedback from multiple sources: support tickets, NPS verbatims, app store reviews, churn interview notes, and survey responses. Separates signal from noise, groups findings by theme with frequency counts, and produces an insight report that identifies the most important problems to solve.

When defining the core use case for a new feature …

/echo-jobs

Takes raw user input in any form: interview transcripts, support tickets, survey responses, or user descriptions and produces structured JTBD statements. Each statement includes the functional job the user is trying to do, the emotional job (how they want to feel while doing it), frequency, intensity, and an underserved score.

Before starting a new research initiative to avoid…

/echo-recon

Scans the repository and linked documents for existing user research: persona documents, interview notes, survey results, JTBD maps, feedback analysis reports, and prior Echo outputs. Summarizes what is currently known about users and flags where knowledge is thin, outdated, or entirely missing.

When the team is building a feature and cannot agr…

/echo-segment

Builds user segments and persona cards from mixed data sources: analytics cohort data, CRM attributes, support ticket patterns, app store review analysis, and qualitative interview signals combined. Each persona includes goals, pain points, behavioral patterns, distinguishing characteristics, and confidence level on each attribute.

When establishing a design system and need the tok…

/form-tokens

Defines a complete design token system: semantic color tokens (primary, surface, text, border, status), spacing scale, typography tokens, shadow and radius values, animation timing tokens, and maps them to CSS custom properties and Tailwind configuration. Bridges design decisions to code so both are consistent.

When adding a new component to a design system and…

/form-component

Produces a component design spec: visual design for all states (default, hover, focus, active, disabled, loading, error, and domain-specific states), interaction behavior and animation, prop definitions with types, accessibility requirements including ARIA attributes, and usage guidelines. Ready for Prism to implement.

When launching a new product that needs a logo. Wh…

/form-logo

Creates a logo through a structured phase-gate process: brand territory research, strategic direction development presenting multiple positioning options, three distinct visual directions for selection, refinement of the chosen direction, and final SVG delivery with usage guidelines including minimum sizes and clear space rules.

When designing a product landing page, marketing s…

/form-web

Designs a landing page or marketing website: section hierarchy and flow, visual layout for each section, content structure and hierarchy, hero section treatment, and conversion flow optimization. Specifies layout, components, and visual treatment at a level of detail Prism can implement without further design clarification.

When designing new screens for a mobile app before…

/form-mobile

Designs iOS and Android screens following platform conventions (iOS Human Interface Guidelines, Material Design): layout, content hierarchy, component usage, navigation patterns, gesture interactions, and platform-specific considerations for each screen. Output is a design spec that Touch can implement.

When setting up email marketing and need templates…

/form-email

Designs HTML email templates for welcome sequences, transactional notifications, newsletters, and drip campaigns. Specifies layout, content block structure, call-to-action placement, image handling, and mobile responsiveness considerations given email client constraints.

When launching a product and need social assets re…

/form-social

Designs social media graphics and ad creatives sized for specific platforms: LinkedIn posts, Instagram stories and feed posts, Twitter cards, Facebook ads, banner ads, and launch announcement assets. Specifies layout, typography treatment, and brand application for each format.

When creating an investor pitch and want the desig…

/form-deck

Designs a pitch deck, investor presentation, or conference slide set: overall narrative structure, slide-by-slide layout recommendations, visual hierarchy approach, data visualization treatment, and design direction for the full set. Specifies design at a level implementable in any presentation tool.

Before a public launch when visual quality matters…

/form-audit

Audits existing UI for visual quality: checks design consistency across pages and components, verifies brand color and typography usage, reviews design system compliance, identifies spacing inconsistencies, and flags visual bugs including misaligned elements, wrong color values, and typography hierarchy violations.

Before launching any experiment where you need to …

/lumen-abtest

Designs an A/B test before it launches: defines the hypothesis in testable form, calculates the required sample size for statistical power, specifies the primary metric and guardrail metrics, sets a minimum detectable effect, determines the required run time, and defines the success and failure criteria clearly.

When adding analytics to a new feature before it s…

/lumen-instrument

Designs an instrumentation plan for a feature or product area: event taxonomy with consistent naming conventions, event property schemas, a tracking plan mapping each user action to an event with its properties, and an implementation checklist. Produces a document both developers and analysts can work from confidently.

When a product team is starting from scratch on me…

/lumen-metrics

Designs a metrics framework: North Star metric with rationale for why it correctly captures product value, input metrics that drive it and can be influenced by product decisions, operational metrics for engineering health, and guardrail metrics that flag regressions. Each metric has a precise definition, an owner, a cadence, and an action trigger.

Before designing new metrics or instrumentation to…

/lumen-recon

Scans existing analytics configuration to map what is currently being measured: event tracking setup and coverage, metric definitions in BI tools, dashboard inventory, and any OKR or goal-tracking documents. Identifies gaps, duplicates, and inconsistencies before designing new measurement.

When establishing the messaging foundation before …

/pitch-message

Produces a full messaging hierarchy: headline with the primary value proposition, subheadline with supporting context, three to five proof points with specific evidence, and a CTA. Designed for consistent use as the source of truth across landing pages, ads, email campaigns, and sales materials.

When a landing page needs copy written, not just a…

/pitch-copy

Writes landing page and marketing copy: hero section with headline, subheadline, and primary CTA; problem and solution block; feature highlights written as benefits; social proof section; and secondary CTA. Written from the perspective of a specific target user who should feel immediately understood when reading it.

When a product or feature is approaching launch an…

/pitch-launch

Creates a GTM launch plan: distribution channels with assigned owners and specific timing, day-1 activities checklist, success metrics for the first 30 days with targets, copy asset checklist (every piece of content that needs to be written before launch), and a simple timeline. Designed to ensure nothing falls through the cracks.

Before writing new positioning or copy to understa…

/pitch-recon

Reads existing marketing materials: landing pages, copy documents, positioning statements, ad copy, email sequences, and sales decks to map the current messaging state. Identifies inconsistencies between surfaces, outdated positioning, and gaps before writing anything new.

Before running any growth experiment to ensure it …

/surge-experiment

Designs a single growth experiment with rigor: states the hypothesis in testable form, specifies the exact metric being tested and how it is measured, establishes a current baseline, calculates minimum detectable effect, determines required run time, and defines explicit kill conditions. One clear, unambiguous design per experiment.

When adding a free tier or self-serve motion to a …

/surge-plg

Designs a product-led growth motion: self-serve onboarding flow that delivers value without a sales rep, freemium tier definition with clear limits that motivate upgrade, upgrade trigger design based on usage signals, in-product virality mechanics, and the expansion revenue path from free to paid. Grounded in what the product actually does.

Before designing new growth initiatives to underst…

/surge-recon

Audits the current growth state: maps existing acquisition channels with estimated contribution and cost, reads onboarding flows step by step, analyzes retention curve shape by cohort, and inventories any prior growth experiments with their outcomes. Produces a factual picture of where growth is coming from and where it is leaking.

When retention is below benchmark and you need to …

/surge-retention

Analyzes retention curves to identify when and why users churn: examines cohort retention by acquisition channel and behavior segment, maps the activities that correlate with retention versus churn, and produces a playbook with specific re-engagement trigger logic, win-back sequence design, and habit loop recommendations.

When designing a landing page structure, ordering …

/draft-landing

Structures landing page information architecture for conversion-optimized layouts. Validates each section against the 'so what?' test and outputs section order with CTA placement markers.

When you need best practices for a specific intera…

/draft-patterns

Reference guide for UX interaction patterns — forms, navigation, loading states, and empty states. Surfaces do/don't tables with severity ratings and code examples.

When evaluating design quality against fundamental…

/form-exam

Theory-backed design audit that names the violated principle, cites the source, and shows the fix. Evaluates across 10 categories from composition to accessibility to AI slop detection.

When a product needs a color palette and the indus…

/form-palette

Generates a full shadcn-compatible color token set (light + dark) matched to product type and industry conventions. Every pair verified against WCAG AA contrast.

When choosing between design directions or visual …

/form-style

Recommends a visual design direction (primary style + fallback) matched to product type and audience. Includes effects, anti-patterns, and an implementation checklist.

When choosing chart types for analytics dashboards…

/lens-chart

Selects the right chart type for analytics and BI dashboards. Evaluates for data density, drill-down capability, real-time support, and accessibility — every chart must answer a decision.

When a landing page needs to be structured around …

/pitch-landing

Structures a launch landing page around positioning, messaging, and conversion. Layers in CTA strategy, social proof placement, and objection handling for a specific audience.

When implementing a chart component or choosing a …

/prism-chart

Selects the right chart type and library for frontend data visualizations. Outputs accessibility grade, interaction level, data volume threshold, color guidance, and an accessibility fallback.

When you need React, Vue, Svelte, Next.js, or Nuxt…

/prism-stack

Framework-specific best practices and component patterns for the detected stack. Detects the major version before applying guidelines — no Vue 2 advice on Vue 3 projects.

Before shipping any UI. Run after design implement…

/proof-design

Design QA audit that scans for visual red flags across typography, color, layout, components, and content. Classifies each finding Critical/Major/Minor and outputs a ship-readiness verdict.

When asked to ship, push to main, or create a PR. …

/relay-ship

End-to-end ship workflow: merge base, run tests, audit coverage, review diff, bump version, commit, push, and create PR. Non-interactive by default — stops only for real judgment calls.

When a landing page is designed for growth: acquis…

/surge-landing

Designs growth-optimized landing pages with activation triggers, funnel structure, and experiment surfaces called out. Built for A/B testing and PLG flows.

When implementing mobile UI or asking about platfo…

/touch-ui

Platform-specific mobile UI guidelines for iOS and Android — touch targets, navigation patterns, gesture conventions, and form design. Always detects platform before outputting rules.

Before any launch or community initiative to under…

/buzz-recon

PR and community reconnaissance: audits current press coverage, social presence, community health, and competitor PR. Produces a factual state-of-the-media picture before planning a launch or community initiative.

When a product or feature is approaching launch an…

/buzz-launch

Designs and executes a coordinated launch plan: Product Hunt, HN Show HN, newsletter coordination, social posts, influencer outreach map, and day-1 activities checklist with a defined success metric.

When a product update, launch, or milestone needs …

/buzz-social

Social media strategy and post drafting: HN Show HN posts, Twitter/X threads, LinkedIn posts, Reddit comments, and developer community content written in voice that passes the authenticity test for each channel.

When reaching out to journalists, podcasters, or n…

/buzz-pitch

Writes media pitches and press releases: journalist outreach emails leading with reader value, podcast pitch scripts, newsletter sponsor pitches, and press release copy written for people who receive hundreds of pitches per day.

When starting a developer community from scratch a…

/buzz-community

Builds and manages open source community: Discord or Slack structure, contributor onboarding flow, ambassador program design, community flywheel mechanics, and GitHub community health files.

Before designing any sales motion or playbook to u…

/deal-recon

Revenue reconnaissance: audits current sales pipeline, deal patterns, ICP definition, and revenue motion to understand what is working and where the constraint is. Produces a factual sales state before any playbook work.

When designing a sales process from scratch for a …

/deal-pipeline

Designs or audits a B2B sales pipeline: stage names, entry and exit criteria, qualification standards, CRM field requirements, and the definition of what moves a deal forward.

When a sales pattern has been established and need…

/deal-playbook

Writes sales playbooks: outbound sequences, discovery call guides with specific questions, objection handling scripts grounded in real objections, and demo frameworks structured around the buyer decision.

When launching a product and pricing has not been …

/deal-pricing

Designs pricing strategy and packaging: tier definitions, value metric selection, enterprise pricing structure, freemium tier limits, and pricing page copy that converts. Grounded in willingness-to-pay and competitive landscape.

When a specific deal has gone quiet and you need t…

/deal-close

Closes a specific deal: diagnoses why it is stalling, writes a tailored proposal for the decision maker, designs the closing sequence, or navigates procurement blockers. Operates on the specific deal, not the general playbook.

Before designing a content strategy to understand …

/ink-recon

Content marketing reconnaissance: audits current content inventory, SEO health, keyword rankings, competitor content gaps, and distribution channels. Produces a factual content state before designing any strategy.

When organic traffic is low and you need a systema…

/ink-seo

SEO strategy and keyword research: builds topic clusters around a core theme, identifies keyword gaps against competitors, performs an on-page audit, and produces a prioritized SEO roadmap with expected traffic impact.

When a blog post needs to be written for a specifi…

/ink-post

Writes a blog post or article from brief to publish-ready: researches the keyword, drafts the post with proper heading structure, adds internal links, and optimizes for the specific search intent the post is targeting.

When content is being produced ad hoc with no edit…

/ink-calendar

Builds a content calendar: editorial plan with topic assignments, publishing cadence that matches team capacity, keyword coverage map, and distribution workflow for each piece from draft to published.

When a successful customer outcome needs to be doc…

/ink-case

Writes customer case studies and success stories: interview guide for the customer conversation, story structure from problem to outcome, and publish-ready case study with specific metrics and a quotable customer voice.

Before designing any CS motion or onboarding impro…

/keep-recon

Customer success reconnaissance: audits current onboarding completion rates, health signal coverage, NRR, churn patterns, and CS motion to understand what is working and where customers are failing. Produces a factual CS state before any playbook work.

When time-to-first-value is too long and customers…

/keep-onboard

Optimizes customer onboarding: maps the activation sequence from signup to first value moment, identifies drop-off points, designs the aha moment, and produces the email sequence that carries customers through setup without human intervention.

When CSMs are reacting to churn instead of predict…

/keep-health

Designs a customer health scoring model: defines the signals that predict expansion and churn, assigns weights based on empirical patterns, sets thresholds for green, yellow, and red health states, and maps action triggers to each state.

When churn is happening and the team has no consis…

/keep-playbook

Writes churn prevention and win-back playbooks: risk intervention sequences triggered by health score drops, save conversation guides for the hardest cancellation conversations, and win-back email campaigns for recently churned customers.

When NRR is below target and expansion is not happ…

/keep-expand

Designs expansion revenue playbooks: upsell triggers based on usage signals, seat expansion sequences for team growth, tier upgrade paths, and cross-sell motions tied to adjacent customer needs.

When starting a new design system and need to reso…

/form-brief

Translates a design brief or plain English description into a concrete DESIGN.md with specific CSS tokens. Resolves palette, accent, typography, display font, layout, mood, density, and constraints into an immutable design contract before any UI work begins.

When a session produced a discovery, fix, or impro…

/pave-contribute

Contributes a session learning back to the upstream tonone repo: scans the conversation for the most reusable insight, asks one clarifying question, and creates a PR with the extracted improvement.

When a security scan needs to be run before a rele…

/warden-scan

Automated SAST and dependency vulnerability scan: runs Semgrep for code vulnerabilities and pip-audit for CVE-matched dependency issues, then writes a structured JSON report with severity classifications.

When launching or formalizing a developer relation…

/buzz-devrel

Designs a developer relations program: community tier structure, ambassador criteria and benefits, DevRel metrics (activation, contribution, retention), and a 90-day DevRel roadmap. Calibrated to team size and current developer engagement baseline.

When planning a Show HN or Ask HN post and want to…

/buzz-hn

Crafts a Hacker News post optimized for traction: title variants following anti-clickbait rules, body text with the story structure that performs on HN, anti-shadowban checklist, and response templates for the top five objection categories the post is likely to receive.

When doing journalist or podcast outreach and need…

/buzz-outreach

Writes personalized media and podcast pitch emails for specific journalists or hosts: researches the target's recent work, identifies the angle most likely to land with that specific person, and produces a short cold pitch calibrated to their beat and tone.

When cold outbound is the primary acquisition chan…

/deal-outreach

Builds a cold outbound sequence: 5-7 touch email and LinkedIn cadence by persona, with subject line variants, opening hooks, value proposition framing, call-to-action design, and follow-up timing. Each message is calibrated to a specific ICP persona and buying trigger.

When a deal has reached the proposal stage and nee…

/deal-proposal

Generates a complete B2B proposal document: executive summary, problem restatement in the buyer's language, solution architecture, pricing and packaging options, ROI model with the buyer's numbers, implementation timeline, and next steps. Built for a specific deal, not as a template.

Before committing engineering or executive time to…

/deal-qualify

MEDDPICC-based deal qualification worksheet: maps each criterion (Metrics, Economic Buyer, Decision Criteria, Decision Process, Identify Pain, Champion, Competition) to the current deal state, scores deal health, identifies the largest gaps, and produces the specific next action to address the critical gap.

Before writing any SEO-targeted content piece to e…

/ink-brief

Generates a content brief for a single piece: target keyword with search intent classification, recommended structure and word count, competitive content gap analysis, internal link targets, external sources to cite, key takeaway framing, and a specific CTA. Produces a brief that a writer can execute without further research.

When starting a content program and need to build …

/ink-cluster

Designs a topic cluster architecture: identifies the pillar page topic, maps 8-15 supporting posts organized by subtopic, defines the internal linking structure between them, and produces a keyword-to-content assignment map that maximizes topical authority for the target domain.

After publishing a high-effort piece that deserves…

/ink-distribute

Produces a distribution plan for a specific piece of content: channel selection rationale, timing and sequencing, platform-specific framing (LinkedIn, Twitter/X, HN, newsletter, community Slacks), repurposing formats (thread, video script, talk abstract), and a 30-day distribution calendar.

When customer health is declining and the CS team …

/keep-churn

Churn risk classification and intervention: scores accounts as CRITICAL, HIGH, or MEDIUM based on behavioral signals and account health, then designs specific intervention sequences for each risk tier including what to send, when, who sends it, and what success looks like.

Before a scheduled QBR with a key account to produ…

/keep-qbr

Generates a Quarterly Business Review document for a specific account: health summary with trend indicators, key wins from the quarter, usage data interpretation, risks and open issues, expansion opportunity framing, and a proposed Q+1 success plan. Built for a specific customer, not a generic template.

When CS headcount cannot scale with the customer b…

/keep-segment

Customer segmentation model for CS coverage: tiers accounts by ARR, health score, and expansion potential, defines coverage model per tier (high-touch, tech-touch, digital-only), sets CS resource allocation ratios, and produces a migration path for accounts moving between tiers.

Before any financial modeling, fundraising, or boa…

/mint-recon

Financial recon: audits burn rate, runway, unit economics health, and cash management. Produces a factual financial state-of-the-company picture before modeling or fundraising work.

When building a financial model from scratch for f…

/mint-model

Builds or audits a 3-statement financial model (P&L, balance sheet, cash flow) with scenario analysis. Produces integrated model with assumptions tab and scenario toggles.

When planning the annual operating budget and need…

/mint-budget

Designs annual operating budget with headcount planning, departmental spend targets, and revenue targets. Produces a budget that aligns resources to strategic priorities.

When runway drops below 12 months and you need a c…

/mint-runway

Calculates current runway and maps all available levers to extend it. Produces a lever analysis with impact estimates and implementation timelines.

When unit economics have not been formally modeled…

/mint-unit

Audits unit economics — LTV, CAC, payback period, gross margin, and cohort contribution margin. Surfaces the metrics that actually matter for the business model.

Before each board meeting when you need a financia…

/mint-board

Produces board financial package: P&L summary, cash position, key operating metrics, and variance vs plan analysis. Formatted for board-level consumption.

When beginning a fundraising process and need to b…

/mint-raise

Prepares fundraising materials: investor financial model, data room structure, cap table analysis, and financial narrative for the pitch. Produces investor-ready financial package.

At month close when you need to produce the financ…

/mint-report

Generates monthly close package with variance analysis, management reporting, and forward-looking indicators. Produces the financial reporting cadence for leadership.

Before any people ops initiative to understand the…

/folk-recon

People operations recon: audits org design, hiring pipeline health, compensation framework, onboarding effectiveness, and performance management maturity. Produces a people org health brief.

When the org has grown past the informal structure…

/folk-org

Designs or reviews org structure: spans of control, reporting line design, headcount plan, and team topology. Produces an org design with rationale for each structural decision.

When opening a new role and need a structured hiri…

/folk-hire

Builds hiring pipeline for a role: job description, sourcing strategy, interview scorecard, and structured evaluation process. Produces a complete hiring process specification.

When the company has no formal comp bands and is m…

/folk-comp

Designs compensation framework: salary bands by level and function, equity grant guidelines, total compensation philosophy, and benchmarking methodology. Produces a comp framework ready for implementation.

When new hires are ramping slowly and there is no …

/folk-onboard

Builds employee onboarding playbook: Day 1 through Week 4 schedule, systems access checklist, role-specific milestones, and 30-60-90 day success criteria. Produces a repeatable onboarding program.

When the company has no formal performance review …

/folk-perf

Designs performance management system: review cycle design, calibration process, career ladder, and goal-setting framework. Produces a performance management program with implementation timeline.

When evaluating which roles can be partially or fu…

/folk-migrate

Human-to-agent migration planning: audits current roles for automation potential, designs transition playbooks for affected functions, and produces a structured migration roadmap. Produces a role-by-role automation assessment with transition sequencing.

When cultural values exist informally but have not…

/folk-culture

Documents and strengthens company culture: values codification, behavioral norms, culture health check, and culture reinforcement playbook. Produces a culture document that is operationally useful rather than decorative.

Before any operations initiative to understand the…

/keel-recon

Operations recon: audits business processes, vendor landscape, compliance posture, OKR execution health, and operational friction points. Produces an ops health brief with prioritized issues.

When a process depends on tribal knowledge and bre…

/keel-process

Documents or redesigns a business process: SOP writing, process mapping, RACI design, and handoff specification. Produces a process document that teams can follow without tribal knowledge.

When vendor contracts are renewing and you need a …

/keel-vendor

Manages vendor relationships: selection scorecard design, contract review checklist, renewal tracking system, and vendor risk assessment. Produces a vendor management framework.

When the company is receiving NDAs and MSAs withou…

/keel-legal

Drafts or reviews legal ops documents: NDA templates, MSA review checklists, SaaS agreement terms analysis, and legal ops process design. Produces legal documents and review frameworks.

When a prospect or customer has asked for a SOC2 r…

/keel-comply

Builds or audits compliance program: SOC2, GDPR, HIPAA gap analysis, control implementation roadmap, and evidence collection process. Produces a compliance program with implementation sequencing.

When the company has never run OKRs formally and n…

/keel-okr

Designs and runs OKR program: objective and key result design, cascade architecture, review cadence, and scoring methodology. Produces an OKR program that drives execution rather than administrative overhead.

When the calendar is full of meetings but decision…

/keel-cadence

Designs meeting cadence: which meetings to run, how frequently, who attends, what decisions each meeting owns, and how information flows between them. Produces a meeting system that replaces ad hoc coordination.

When operational costs are growing faster than rev…

/keel-audit

Operational efficiency audit: scans for waste, redundancy, and friction across business processes, tools, vendors, and workflows. Produces a prioritized list of efficiency improvements with effort and impact estimates.

Before any support operations initiative to unders…

/brace-recon

Support operations recon: audits ticket volume patterns, SLA compliance, CSAT trends, and knowledge base coverage gaps. Produces a support health brief with prioritized structural issues.

When all tickets go into one queue regardless of s…

/brace-triage

Designs ticket triage system: routing rules by category and severity, priority tag taxonomy, queue structure, and assignment logic. Produces a triage specification ready to implement in any support platform.

When the same questions appear in tickets repeated…

/brace-kb

Builds or audits knowledge base: coverage gap analysis, deflection rate diagnosis, article structure standards, and maintenance workflow. Produces a KB program that reduces ticket volume by keeping documentation current.

When customers are asking for SLA commitments and …

/brace-sla

Designs SLA framework: response time targets by severity tier, resolution time targets, tier definitions, and breach escalation paths. Produces a defensible SLA document suitable for customer contracts.

When complex tickets bounce between support tiers …

/brace-escalate

Designs escalation path: Tier 1 to Tier 2 to Engineering handoff criteria, escalation SLAs, context transfer templates, and de-escalation thresholds. Produces an escalation system that prevents tickets from stalling at tier boundaries.

When new customers frequently submit setup tickets…

/brace-onboard

Designs support onboarding flow: first-contact experience design, product setup verification checklist, initial health check, and early warning triggers. Produces an onboarding support protocol that prevents early churn.

When support performance is measured by ticket cou…

/brace-metrics

Designs support metrics dashboard: CSAT, first response time (FRT), time to resolution (TTR), ticket deflection rate, and trend analysis. Produces a metrics framework with leading and lagging indicators.