rahulnakmol-skills
AI-native skills across six groups: shared doctrine, the software delivery lifecycle, product and transformation, branding, writing, and productivity
Not in a marketplace yet, so it loads for one session with --plugin-dir.
Plugin rahulnakmol-skills in akmolsinc/skills. Commands read from the repo's manifest — 25 days ago.
What it installs
Skills (64)
core-doctrineDoctrine documents of the core group in tqnonline/skills, packaged for Amp as `COVERAGE.md`, `GRILL.md`, `TRACE.md`, `VALUE.md`, and `VERIFICATION.md`. Use when a skill installed from that catalog cites one of these files, or to read the rules the core skills work under. Skills in every group cite t
developer-doctrineDoctrine documents of the developer group in tqnonline/skills, packaged for Amp as `DDDD.md`. Use when a skill installed from that catalog cites one of these files, or to read the rules the developer skills work under. Skills in the developer group cite these documents by bare filename.
pm-doctrineDoctrine documents of the pm group in tqnonline/skills, packaged for Amp as `AGENT-OWNERSHIP.md`, `DDDD.md`, `GATES.md`, `HATS.md`, `INITIATIVE-REPO.md`, `RESEARCH-AGENTS.md`, and `VISUALS.md`. Use when a skill installed from that catalog cites one of these files, or to read the rules the pm skills
tqnRouter for the tqnonline/skills catalog when it runs inside Amp. States where a skill's doctrine citation resolves after an Amp install, which Amp tool to prefer where one overlaps a catalog skill (Librarian, Finder, Oracle, threads, modes, schedules, charts), and how the catalog's agent and human n
ai-brandingUser-invoked AI branding skill. Use for an Anthropic-inspired editorial system by default, or set provider openai for a separate monochrome profile with open-source fonts.
branding-systemModel-invoked branding engine. Use when a theme skill supplies a validated profile and an artifact needs one coherent verbal, visual, accessibility, physical, or motion system.
catppuccin-brandingUser-invoked theme skill. Use for expressive pastel branding across digital, print, environmental, and motion surfaces with accessible Catppuccin light and dark roles.
everforest-brandingUser-invoked theme skill. Use for calm, natural, editorial branding across digital, print, environmental, and motion surfaces with an accessible Everforest profile.
exhibitBuild one self-contained, interactive HTML page that walks a reader through a topic as a journey, in the colors, type, motion, and voice of a named brand. User-invoked. Use when someone asks to "explain this topic as an interactive page", "make an explorable explainer", "visualize how this works in
gruvbox-brandingUser-invoked theme skill. Use for warm retro-modern branding across digital, print, environmental, and motion surfaces with accessible Gruvbox light and dark roles.
nord-brandingUser-invoked theme skill. Use for cool, spacious branding across digital, print, environmental, and motion surfaces with accessible Nord light and dark roles.
pressRender an approved markdown document into a branded, self-contained HTML page with a clickable table of contents, and a PDF — a product requirements document, board pack, decision record, research summary, or runbook. User-invoked. Triggers include "turn this into a branded PDF", "render the doc for
solarized-brandingUser-invoked theme skill. Use for disciplined blue-and-gold branding across digital, print, environmental, and motion surfaces with accessible Solarized light and dark roles.
gritUser-invoked completion discipline for substantial AI-agent work, backed by runnable gates. Use to write the acceptance-gate ledger before implementation, choose a verification depth, run the four passes, and audit met, unmet, and abandoned gates before done is declared.
researchModel-invoked source investigation. Use when a question needs an answer from primary sources — official documentation, a specification, or the source code itself — and the finding should be written to a cited file in the repository so a later session reads it instead of repeating the search. Where t
retroUser-invoked improvement loop, the sixth layer of the harness. Use to read accumulated run traces, find the failure patterns that repeat across three or more runs, and propose changes to the skills themselves as a pull request a human reviews before it merges.
architectCross-cutting technical design, ADRs, decomposition. Use at design gate or from recon brief.
ask-fdeUser-invoked skill router. Maps intent to the correct developer or branding skill. Use when unsure which skill to run.
atlasBuilds evidence-grounded reference architecture documentation for Azure, AWS, or GCP, with official-icon SVG diagrams, branded HTML/PDF, and independent quality judgment. User-invoked. Use for architecture packs from repositories or requirements, or cumulative architecture revisions.
conductModel-invoked routing for work that builds or changes software. Decides whether delivery runs as a loop, a graph, or a hybrid, assigns a model to each node, and maps the result onto harness adapters. Use before implementation spanning several workstreams, verifiers, or context windows begins. For ch
debugModel-invoked diagnosis loop for hard bugs and performance regressions. Use when a defect resists a quick fix, when a failure is intermittent, or when a measurement has moved the wrong way and the cause is not yet known.
deliverModel-invoked charter for the delivery lane — continuous integration and delivery, supply chain, and release readiness. Use at the delivery gate to check a repository against the readiness checklist, ship a multi-concern change as a stack of dependency-ordered pull requests, resolve a merge or rebas
glossaryModel-invoked construction of a project's shared domain vocabulary. Use when the same concept carries several names across code and conversation, or before a new agent or contributor needs the project's language written down.
impactUser-invoked idea-to-PRD pipeline with grill loop, value probing, and backlog handoff. Use for greenfield or brownfield inception before slice and raise.
model-routingModel-invoked lookup of tier and role assignments from the canonical registry. Serves every group in this repository, developer and pm alike. Use when selecting a worker, verifier, or orchestrator model for a task node. Where the tool exposes effort modes rather than model identifiers, bind the tier
operateCharter for a system after it is built — reliability, quality, and maintenance in one lane. Covers service-level objectives, error budgets, instrumentation, dashboards, runbooks, on-call and incident readiness; quality assurance of the SPEC-TS ledger, test gaps, coverage floors, and technical-debt f
prototypeModel-invoked throwaway prototype that answers one design question. Use when a state-model, logic, or interface choice needs running evidence before it is committed to, and when the build can be deleted or explicitly graduated once the answer arrives.
raiseModel-invoked publish of sliced backlog to GitHub, Linear, or Azure DevOps with pickup-protocol labels. Use after slice output is ready.
reconModel-invoked brownfield codebase brief via signal-first archetype triage. Read-only; loads ≤3 pattern cards. Use before impact or architect on existing estates. Where the tool has its own code search or codebase-explanation subagent, gather signal through it; this skill structures the brief.
refactorUser-invoked survey of a codebase for places where a module could hide more complexity behind a smaller interface, ranked and handed to a human to choose from before any structural change begins. Use when the goal is to reduce the complexity callers must carry, not to fix a defect.
responsible-ai-governanceResponsible-AI and regulated-industry governance overlay. Use when work touches a regulated industry (financial services, healthcare, public sector) or an AI system that makes or influences consequential decisions, or when the user asks about audit, explainability, model risk, residency, NCUA, OSFI,
safeguardSecurity assessment and hardening charter. Use at secure DevOps gate (replaces generic security-review).
sdlcUser-invoked full gated SDLC loop with SPEC-TS ledger, human gates, and verifier challenge. Use when executing a work item or feature through design, build, secure, and release gates.
shakedownUser-invoked PR sandbox build, test, and review. Use before merge or when validating a contributor PR in isolation. Where the tool provides an isolated sandbox of its own, build and run the branch there; this skill supplies the checklist and the verdict format.
sliceModel-invoked PRD decomposition into epics, features, stories, and operability items conforming to WORK-ITEM-CONTRACT. Use after impact PRD sign-off.
tddModel-invoked test-first discipline. Use to build a feature or fix a bug one vertical slice at a time, writing a failing test that fails for the right reason before the implementation exists, and refactoring only under green.
triageUser-invoked triage of incoming issues and external pull requests. Use to categorize a report, confirm it reproduces, and leave an agent-ready brief on the issue so a human can decide what the project accepts.
update-modelsUser-invoked model registry curation. Research provider catalogs, propose updates to models.md and adapters, open a reviewable PR. Serves every group in this repository, developer and pm alike — the registry it curates is not developer-only. Use on schedule or when models deprecate. Where a tool cho
wizardModel-invoked generator for an interactive shell wizard that walks a person through steps only a person can perform. Use when work stalls on manual provisioning, credential setup, an unfamiliar third-party dashboard, or a one-off migration or cutover.
arrangeModel-invoked routing for product and program inquiry, where the work is questions and evidence rather than code. Decides whether a discovery, business case, target operating model, or PRD effort runs as a grill loop, a parallel fan of independent research, or a hybrid. Use before a multi-round effo
ask-pmUser-invoked skill router for the pm group. Maps intent to the correct pm skill. Use when unsure which pm skill to run.
carveModel-invoked epic decomposition from a Business Understanding Document or a TOM's maturity gaps. Use to carve DIVE-tested epics into a manifest before PRDs are drafted, one epic at a time.
caseUser-invoked business case builder. Use to weigh at least two real options plus do-nothing, cost each including the agent fleet's own cost, and prepare the case the sponsor approves at the Investment gate.
chartUser-invoked charting of work too large for one agent session. Turns an initiative into a chart of decision tickets on the tracker, each sized to one session, resolved one decision at a time by a PM and an agent team. Use when an initiative is too big, too uncertain, or too cross-cutting to run the
constitutionUser-invoked Product Constitution author and reviewer — principles, positioning, and the prioritization framework every downstream pm skill reads for alignment. Use to create, evolve, or review the practice or an initiative's constitution.
discoverUser-invoked business problem discovery and root-cause analysis for product and program managers. Use to turn a raw problem, meeting notes, or a vague opportunity into a structured, stakeholder-grounded analysis before solution design.
grillUser-invoked interrogation loop for PM artifacts — plain, grill-with-docs, and provoke modes. Use to pressure-test an analysis, manifest, case, or PRD before it advances through a gate.
mapModel-invoked persona mapping, process-flow diagramming, and Business Understanding Document assembly. Use after discover hands off a completed analysis, to produce personas, current/target Mermaid flows, and the document the Framing gate signs.
prd-draftUser-invoked PRD drafting and structural validation from an approved epic manifest. Use to generate one PRD per epic with INVEST-compliant user stories, Given-When-Then acceptance criteria, success metrics, and a grill pass, then to run the nine-check structural validation that ends in a PASS, PASS
prd-reviewUser-invoked PRD quality reviewer using the 11-Star Experience Framework. Use to score a PRD across seven weighted dimensions and map its features to star levels, informing but never replacing the human decision at the Quality gate.
raidUser-invoked RAID register maintainer — Risks, Assumptions, Issues, Dependencies. Use to keep the four registers current and to generate the two-minute sponsor status narrative directly from them.
realizeUser-invoked benefits realization tracker. Use to maintain the benefits register against case and PRD projections, track leading indicators and actuals, roll up to the north star, and loop variance back into new work.
reportUser-invoked leadership report generator. Use to produce the 4Ps leadership pack — Progress, Problems, Priorities, Perspective — at weekly, fortnightly, quarterly, half-yearly, or yearly cadence, sourced from the registers and the live pipeline.
roadmapUser-invoked roadmapping and PI planning. Use to sequence initiatives by outcome linkage and dependency across now/next/later horizons, and to run Program Increment planning where the practice uses it.
tom-architectUser-invoked Target Operating Model architect for enterprise transformation. Use to decompose business operations into L1-L4 processes, assess maturity, design organization and RACI, and map capabilities to an enterprise platform stack.
briefUser-invoked authoring of the rules, definitions, and boundaries a team's AI agents read. Breaks a working agreement into always-loaded rules, a shared glossary, and on-demand references, then places each part in the file the tool actually loads — Claude Code, Claude Desktop, Cowork, OpenCode, Codex
handoffUser-invoked compaction of an in-flight run into a handoff document a cold reader can pick up. Use when a session ends with work unfinished, when the context window is nearly full, or when work passes to another agent or person who was not present for it. Where the tool keeps session history across
questionnaireUser-invoked construction of a written questionnaire for the one person who can settle a decision the run cannot settle alone. Use when work is blocked on knowledge a named individual holds, and the answers will be filled in asynchronously or worked through together in a meeting.
spotlightTurns established context into the smallest source-grounded visual that helps a human understand or decide. Use when asked to visualize a flow, structure, relationship, comparison, or proposed change. Where the tool renders charts or diagrams inline, render through it; this skill chooses the one vie
teachUser-invoked teaching of a skill or concept across several sessions, which uses the working directory as a stateful workspace so the level established, the exercises set, and the sticking points survive between sessions — use it when the goal is for the learner to hold the knowledge afterward rather
wait-whatUser-invoked re-pitch of a message that did not land, which diagnoses why the explanation failed and rebuilds it from what the reader already understands — use it the moment you read something and do not follow it.
draftUser-invoked render stage of the writing sequence — use it once an outline of beats exists to write prose one beat at a time and record which beat each paragraph serves, so the draft's coverage of the outline is auditable rather than felt.
freewriteUser-invoked explore stage of the writing sequence — use it at the start of a piece, before any structure exists, to mine raw fragments from the author while keeping generation strictly separate from judgment.
outlineUser-invoked structure stage of the writing sequence — use it after freewriting to assemble raw fragments into an ordered sequence of beats, grounding every term before a beat leans on it and cutting material that serves no beat.
Maintain rahulnakmol-skills?
This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.