all publishers

spencermarx

@spencermarx source repo

47 published skills

  1. ▌
    Write Comments · spencermarx
    Write code comments the Ousterhout way AS you author or change code: comment only the non-obvious, write interface and header comments first, and let clear names carry the rest. Use while implementing a function, class or module, adding or editing a comment, writing a doc comment, JSDoc block or docstring, or deciding "should this have a comment, and what should it say". Triggers: "add a comment", "document this function", "write the header/docstring/JSDoc", "should I comment this", "comment this code". Do NOT use to audit comments that already exist (use `clean-comments`).
    0 installs
  2. ▌
    Writing For Agents · spencermarx bundle
    The craft of the prose agents read: skill descriptions and bodies, instructions files (AGENTS.md, CLAUDE.md), rule files and reference docs; context pointers, progressive disclosure, leading words and pruning. Use whenever wording or rewording any of those, deciding what belongs in always-loaded context versus behind a pointer, pruning an instructions file that has grown, or asking whether a line is a no-op. Do NOT use to build, split, port or test a skill end to end (use `skill-builder`, which loads this), to measure whether one fires (`harness-eval`), to state or route a convention (`add-standard`, `standards`), or for prose a person reads: a working surface (`conveying-clearly`), a byline piece (`draft-argued-piece`), or marketing copy (`seven-copy-critics`).
    0 installs
  3. ▌
    Draft Argued Piece · spencermarx
    Drafts a piece block by block from existing material: settles the thesis, then grows the argument in the order the reader needs it. Use when a pile exists and someone says write the blog post, draft the essay, turn these notes into an article, the newsletter issue, whitepaper, guide, case study or talk script. Do NOT use with nothing written (use `gather-argument-material`), for a cold email or landing page (use `draft-conversion-copy`), or to settle a decision (use `grill-me`).
    0 installs
  4. ▌
    Seven Copy Critics · spencermarx bundle
    Adversarial copy review that stress-tests marketing copy through seven hostile reader personas plus a byline-voice check, each at three failure layers. Catches weak hooks, generic phrasing, unsubstantiated claims, and abstraction. Produces a failure report and a tightened rebuild. Use before sending cold email, publishing a page, or scheduling a post. Reviews drafts only: never use it to write copy. Drafting an essay or post is `draft-argued-piece`, a cold email or landing page `draft-conversion-copy`; reviewing either is here. `humanize` strips AI tells first. Do NOT use to review a whole campaign brief (use `campaign-brief`) or visual creative (use `brand-review`).
    0 installs
  5. ▌
    Ocr Review Loop · spencermarx bundle
    Drive a pull request to an approved code review by looping Open Code Review's review and address steps until the verdict is APPROVE, posting every round to the PR as comments. Use when someone says review and address in a loop, iterate review until approved, auto review-and-fix this PR, loop the OCR review and address, or take a changeset all the way to a clean APPROVE. Needs a feature branch with an open GitHub PR. Do NOT use for a single review with no fixing (run OCR's review directly), or to address one existing review (run OCR's address directly).
    0 installs
  6. ▌
    Rice Prioritization · spencermarx bundle
    Score and rank a list of candidate features or initiatives with the RICE model (Reach x Impact x Confidence / Effort), using a deterministic calculator, then pressure-test the scores and write a ranked, sensitivity-checked analysis ending in a decision block for the user. Use whenever a backlog or shortlist needs ordering against one goal, a roadmap ordering debate needs numbers, or someone asks to compare initiatives on impact per effort or to RICE-score something. Do NOT use for a go/no-go on one undrafted idea (use `strategic-brief`), for a leadership review of one drafted plan (use `ceo-review`), for a marketing campaign (use `campaign-brief`), or to make the final call, which stays with the user.
    0 installs
  7. ▌
    Draft Conversion Copy · spencermarx
    Drafts conversion copy, where the piece is an ask not an argument: cold outbound email and sequences, landing or services page copy, ads, a short promotional social post. Use when someone says write the cold email, the outreach, the landing page, the ad, or the launch post. Do NOT use for a piece that argues in order (use `draft-argued-piece`, or `gather-argument-material` when nothing is written yet), or to review drafted copy (use `seven-copy-critics`).
    0 installs
  8. ▌
    Ubiquitous Language · spencermarx
    Create or edit a repository's glossary file (its name, place, sections and entry format) or the glossary map that indexes one file per bounded context. Use whenever a glossary file must be scaffolded, written to, restructured or checked: set up the glossary, fix its format, add or reword an entry, write or update the glossary map, split one context into several, "where do domain terms live", a map row pointing at a missing file, or when another skill has a settled term to write down. Owns the format and the file. Do NOT use to decide which terms matter, to pin down or settle a ubiquitous language, or to judge whether the language is right (use `domain-modeling`), or to record a decision (use `decision-record`).
    0 installs
  9. ▌
    Gather Argument Material · spencermarx
    Mines raw material for a piece of writing: interviews the author relentlessly, appending every fragment to one working file and imposing no outline. Use whenever someone wants an essay, newsletter, post, guide or talk and has nothing written yet, is still adding to a pile, or wants to think out loud first. Do NOT use once the material is there and the piece must be drafted (use `draft-argued-piece`), for conversion copy such as a cold email (use `draft-conversion-copy`), or to settle a choice (use `grill-me`).
    0 installs
  10. ▌
    First Principles Investigation · spencermarx bundle
    Evidence-first root-cause investigation: competing hypotheses, each tested against the actual system, until exactly one survives and the fix is proven. Use whenever something behaves in a way nobody can explain, a fix did not hold, the obvious cause is already ruled out, or the question is how a system really works: "why is CI flaky", "get to the bottom of this 500", "why does the deploy drift from main", "root-cause this failure", "debug this properly", "this makes no sense, figure out what's going on", "trace where this value comes from", even when the user only describes a confusing behavior. Prefer it over a domain skill while the cause is still unknown. Do NOT use when the cause is known and only the fix is in question, to stress-test a plan (use `grill-me`), to review a diff, or for a one-line lookup you can answer directly.
    0 installs
  11. ▌
    Get Brand Kit · spencermarx bundle
    Look up a brand value BEFORE making anything that must look or sound on-brand: a slide, an image prompt, an infographic, a social asset, a landing page, an email graphic. The repository's single source of truth for colours, typography, logos, the visual design system and voice; other skills load it by name. Ships a placeholder brand to replace with your own. Do NOT use to review copy (use `seven-copy-critics`), to audit a finished asset (use `brand-review`), or to redesign a product interface (use `design-expert`, which loads this for its values).
    0 installs
  12. ▌
    Prototype · spencermarx bundle
    Builds a throwaway prototype that answers one design question: a single-file HTML demo driving a state model, or radically different UI variants on one route, then folds only the validated decision into real code. Use whenever someone wants to try something out before committing to it: "prototype this", "mock it up", "spike it", "does this logic feel right", "will this state machine hold up", "try a few layouts", "show me options before we build", "a throwaway demo a non-developer can drive", even when they don't say "prototype". Do NOT use to build or polish the real thing for keeps, or to settle a choice by interview (use `grill-me`).
    0 installs
  13. ▌
    Session Retro · spencermarx bundle
    Retrospective on one agent session: reads the session back (a bundled digest for Claude Code's JSONL transcripts, any other harness's log, or the agent's own context), finds what the repository made hard, and proposes fixes by severity, each for the layer that owns it. Use whenever someone says run a retro, retro this session, what slowed that run down, why did that take so many turns, or make this repo easier for the next agent. Do NOT use to measure whether a skill fires (use `harness-eval`), to write the fix itself (use `skill-builder` for a skill, `add-standard` or `standards` for a convention, or `writing-for-agents` to reword an instructions file), or for a retrospective on a project or a team's week rather than one session.
    0 installs
  14. ▌
    Skill Builder · spencermarx bundle
    Author, refactor, split or port an agent skill (a folder with a SKILL.md) so it does one thing, composes with other skills, works in any harness that reads SKILL.md, and fires when it should. Use whenever someone says make this a skill, turn this workflow into a skill, write a skill for X, refactor or split this skill, port this skill to another agent, this skill is not firing, or asks whether something should be a skill, an instructions-file note, a standard, a subagent or a command. Do NOT use to measure whether a skill fires (use `harness-eval`), to state or route a convention (use `add-standard` or `standards`), or for the craft of agent-facing prose in general (use `writing-for-agents`).
    0 installs
  15. ▌
    Brand Review · spencermarx
    Design audit of finished marketing creative against the brand: landing pages, social graphics, email templates, ad creatives. Scores each asset 0 to 10 on seven dimensions, separates mechanical fixes from design judgment calls, and returns a scorecard plus a prioritized fix list. Use before publishing any visual asset, or on is this on-brand, audit this creative, or review this design before it goes out. Do NOT use for copy quality (use `seven-copy-critics`), for the campaign's strategy (use `campaign-brief`), to look up brand values while making an asset (use `get-brand-kit`), or to design or fix a product interface (use `design-expert`).
    0 installs
  16. ▌
    Conveying Clearly · spencermarx bundle
    The contract for a working surface a colleague reads once and acts on: a statement, a question or option set, a verdict, a ticket or PR comment, a status update. Use BEFORE composing one; `wayfinder` loads it automatically. Also trigger the moment someone says please clarify, I don't understand the question, wait what, or I'm not following where you've got to. Do NOT use for prose published under a byline (an essay, post or newsletter), or to compact a session for a successor (use `handoff`).
    0 installs
  17. ▌
    Arch Board · spencermarx bundle
    Convenes a board of whatever personas a design challenge needs (for example Martin Fowler, Rich Hickey, a security lens) that investigates it independently, debates the cruxes and returns a cited verdict with dissents kept. Use whenever a design fork is hard or cross-cutting, or on convene a board, get a panel to weigh in, deliberate the trade-offs, or what would <expert> think about this. Do NOT use to review an existing diff (`ocr-review-loop`), to root-cause a failure (`first-principles-investigation`), to interview the user about a plan (`grill-with-docs`), or for a change with one obvious approach.
    0 installs
  18. ▌
    Campaign Brief · spencermarx
    Adversarial pre-launch review of a marketing campaign brief: challenges it on six dimensions (audience specificity, dream outcome, demand evidence, differentiation, measurability, message-audience fit), rewrites the brief, and issues a GO, NO-GO or CONDITIONAL GO. Use before an outbound sequence, a paid ad test, a content push or a co-marketing deal ships, or whenever someone asks is this campaign ready. Do NOT use to write the copy (use `draft-conversion-copy`), to review drafted copy line by line (use `seven-copy-critics`), to audit finished creative (use `brand-review`), or to judge a plan that is not a campaign (use `ceo-review` or `strategic-brief`).
    0 installs
  19. ▌
    Decision Record · spencermarx
    Record a settled choice and its reasoning as a decision record (ADR) so nobody relitigates it. PROPOSE one for a precedent-setting, hard-to-reverse or surprising decision, then stop: only the user's yes creates one. Trigger on "write an ADR", "add an ADR", "record this decision", "write this down", "document this decision", "supersede that decision", "amend that record", "that ADR is wrong", "why did we choose", or an "X vs Y" choice that just settled. Do NOT use for a choice you could undo in an afternoon, to park work for later, or for work not yet decided (use `wayfinder` or `grill-me`).
    0 installs
  20. ▌
    Domain Modeling · spencermarx
    Build and sharpen a project's domain model while designing. Use when the user wants to pin down a domain's vocabulary or settle what a term means, uses a vague or overloaded word, notices a term being used two ways, says the code and the language disagree, or when a significant design decision surfaces that may deserve a record, and when another skill needs to maintain the domain model. Owns the judgment of which terms matter and whether the language is right. Do NOT use to scaffold, format or restructure the glossary file (use `ubiquitous-language`), merely to read the glossary for vocabulary, or to author the decision record itself (use `decision-record`).
    0 installs
  21. ▌
    Grill With Docs · spencermarx
    Relentlessly interview the user to settle an ambiguous plan, design or request, then capture what it settled as glossary entries and decision records. Use whenever a request carries a choice you cannot settle alone and one question will not clear it, before building anything, and when the outcome should be written down. Do NOT use when a single question would do, or when nothing needs recording afterwards (use `grill-me`).
    0 installs
  22. ▌
    Strategic Brief · spencermarx bundle
    Demand validation before committing to an idea that is not yet drafted: a product feature, a strategic initiative, or a significant marketing investment. Runs a six-question demand diagnostic, writes an initiative brief, challenges it adversarially on demand evidence, scope creep, feasibility and success criteria, then issues a GO, NO-GO or CONDITIONAL GO. Use before committing more than about two days of work, when a request arrives without demand evidence or success criteria, or on is this worth doing, should we build this, or write the brief. Do NOT use to review a plan already drafted (use `ceo-review`), to rank several candidates (use `rice-prioritization`), to review a drafted campaign brief before launch (use `campaign-brief`), or to audit finished creative (use `brand-review`).
    0 installs
  23. ▌
    Create HTML Slides · spencermarx bundle
    Build a branded slide deck as one self-contained HTML file: 16:9, editorial, animated by default, with an optional PDF export. Use for a sales or demo deck, an internal presentation, a pitch, a workshop or conference talk, or to improve an existing HTML deck. Do NOT use for a report or visual write-up read on screen (use `create-report`), a product interface (use `design-expert`), a throwaway UI mock (use `prototype`), or a deck others must edit natively in a slides app.
    0 installs
  24. ▌
    Setup Aclarify Skills · spencermarx bundle
    Configure this repository for the Aclarify skills: where issues, the glossary, decision records and scratch files live, then check and install the tools some skills need, such as the ocr CLI for code review. Run once per repository before first use, and again with `check` to re-verify the tooling.
    0 installs
  25. ▌
    Clean Comments · spencermarx
    Audit and repair EXISTING code comments so each says what the code cannot (intent, rationale, invariants, units, gotchas) and none restates it: delete restatement, banners and commented-out code, fix stale or wrong comments, sharpen bloated ones. Use whenever the user asks to clean up, audit, review or fix comments, remove redundant or obvious comments, tidy doc comments or docstrings, strip commented-out code, "make comments follow Ousterhout", or improve comment quality across a file, a directory, or a changeset (staged diff, commit range, PR), even if they only say "the comments need work". Do NOT use for comments on code you are writing right now (use `write-comments`).
    0 installs
  26. ▌
    Humanize · spencermarx bundle
    Rewrites existing text so it reads as a person wrote it: profiles cadence and AI tells, rewrites by tier, then checks it blind. Use whenever someone says humanize this, make it sound human, make it sound like me, match my voice, de-AI it, this reads like ChatGPT, or a detector flagged it. Do NOT use to draft a piece from nothing (use `gather-argument-material`), to review copy for conversion (use `seven-copy-critics`, after this), or for an internal working surface (use `conveying-clearly`).
    0 installs
  27. ▌
    Standards · spencermarx bundle
    Route, change, move or retire a convention in a repository's conventions folder (its written house rules), keeping every pointer at it in step: the instructions file, the folder's index, and any harness rule files that load it when a matching file is touched. Use whenever you edit a convention, add or change a rule file that points at one, rename a heading something links to, or someone asks where does this belong, is this routed, or what enforces this. Always use before deleting a script or check that enforces a convention in code. A brand new convention starts at `add-standard`, which hands back here. Do NOT use to author a skill (use `skill-builder`) or to measure whether a skill fires (use `harness-eval`).
    0 installs
  28. ▌
    Grill Me · spencermarx
    Relentlessly interview the user about a plan, decision, design or idea, one round of questions at a time, until every branch is settled and nothing is silently assumed. Use when the user wants to stress-test a plan or their thinking before building, when a choice cannot be settled alone and one question will not clear it, or on any 'grill' trigger phrase ("grill me", "grill this plan"). Do NOT use when what the interview settles must be written down as glossary entries or decision records (use `grill-with-docs`), when a single question would do, or to gather raw material for a piece of writing.
    0 installs
  29. ▌
    Whiteboard · spencermarx bundle
    Stand up a live, shared tldraw whiteboard the user and agent sessions draw on together in real time. Use when someone types /whiteboard or says draw with me, sketch together, live or collaborative whiteboard, whiteboard it with the agents, or real-time diagramming. Not for a static diagram saved into a document and never revisited: draw that directly in the document's own format.
    0 installs
  30. ▌
    Create Report · spencermarx bundle
    Assemble a rich visual HTML report, rendered from a markdown record as a self-contained local file, with an optional publish as a shareable page. Use WHENEVER work deserves a report, briefing, or visual write-up rather than chat prose, and whenever another skill needs its results rendered and delivered. Do NOT use to write prose a stranger reads straight through (an essay, post, guide or case study), or for an interactive page or app.
    0 installs
  31. ▌
    Triage · spencermarx bundle
    Move issues, and external pull requests where the repository treats them as requests, through a small state machine of triage roles - categorise, check it is not already built, underway or rejected, verify the claim, grill it into shape, then write the agent-ready brief. Use to triage the tracker, evaluate a named issue or PR, see what needs attention, see what is ready for an agent, or set a triage state. Writes to the issue tracker (labels, comments, closures). Do NOT use to deliberate a hard design fork (`arch-board`), to root-cause a failure outside triage (`first-principles-investigation`), or to chart a multi-session effort (`wayfinder`).
    0 installs
  32. ▌
    Ceo Review · spencermarx
    Executive review of a drafted plan, proposal or roadmap item: commits to one stance (EXPAND, CHERRY-PICK, HOLD or CUT), judges the plan on eight business dimensions, and returns stance-specific scope changes plus at most five ranked open questions. Use when a plan is on the table and needs a hard leadership read: quarterly roadmap priorities, a significant new bet, a pivot, or a large resourcing call. Do NOT use to test whether demand exists before anything is drafted (use `strategic-brief`), to rank several initiatives against each other (use `rice-prioritization`), to review a marketing campaign (use `campaign-brief`), or to audit finished creative (use `brand-review`).
    0 installs
  33. ▌
    Watch CI · spencermarx bundle
    Watch a pull request's or branch's CI run, or a deploy run, to its conclusion and report the verdict; on a failure, triage it and, once the user agrees, fix, push and watch again until green. Use whenever someone wants to monitor, watch, track or wait on CI, a pipeline, a workflow run, a PR's checks or a staging or production deploy: "watch ci", "is CI green yet", "did the run pass", "wait for the deploy to finish", "tell me when the build is done", "get this PR's checks green". Especially when a background or orchestrated agent must block on a run: the tracked wait here is what keeps it from silently stalling. GitHub Actions through the `gh` CLI by default; other CI providers by the same method. Do NOT use to kick off a deploy from scratch, or to triage test failures that are not tied to a CI run.
    0 installs
  34. ▌
    Add Standard · spencermarx bundle
    Turn one convention someone said out loud into a stated, routed standard in the repository's conventions folder: triages that it is a standard at all, drafts the section with the user, then hands the siting and routing to `standards`. Use whenever someone says we should always X, we should never Y, write this convention down, make this a rule, add this to our standards, or that belongs in the house rules. Do NOT use to change, move, retire or audit an existing standard (use `standards`), to author a skill (use `skill-builder`), or for the craft of agent-facing prose (use `writing-for-agents`).
    0 installs
  35. ▌
    Harness Eval · spencermarx bundle
    Measure whether a repository's skills fire when they should: routing evals that show which skill a prompt actually triggers in Claude Code, plus harness-neutral static checks (skill inventory, description cap, Claude Code rule glob reach). Use whenever someone says run the evals, did that skill trigger, why did the wrong skill fire, is this description any good, write routing cases for this skill, or check the harness. Do NOT use to write or rewrite the skill under test (measure here, then edit it), or to review a code change.
    0 installs
  36. ▌
    Design Expert · spencermarx bundle
    Design, redesign, enhance, or fix a product interface with senior UX/UI craft: audit the current state by severity, set the design intent, then implement layout, type, color, interaction, and responsive polish. Use on redesign this component, improve this UI, fix the visual hierarchy, make this feel premium, or this screen looks off. Do NOT use to audit marketing creative (use `brand-review`), to look up brand values for a marketing asset (use `get-brand-kit`), to build a deck (use `create-html-slides`), or to explore throwaway variants before building (use `prototype`).
    0 installs
  37. ▌
    Handoff · spencermarx bundle
    Compact the current session into a handoff document: the load-bearing references, decisions and resume coordinates, none of the transcript noise, so a fresh agent carries it on. Use whenever someone says hand this off, write a handoff, I am running out of context, note where to resume, or pick this up in a new session. Do NOT use to word one message well (use `conveying-clearly`), to open a plain scratch file (use `scratchpad`), or to record a settled decision (use `decision-record`).
    0 installs
  38. ▌
    Teach · spencermarx bundle
    Teach the user a new skill or concept over multiple sessions, grounded in why they want to learn it, with lessons, reference sheets and checks for understanding.
    0 installs
  39. ▌
    Scratchpad · spencermarx bundle
    Create, find, and reclaim ephemeral, git-ignored working artifacts in the repository's one scratch folder (`.scratchpad/` by default): handoffs, research dumps, generated reports, debug output. Reach for it WHENEVER you are about to write a temporary file that must not be committed, instead of an ad-hoc `tmp/` or a stray `notes.md`. Also to list prior entries or clean up old ones. Do NOT use for durable tracked artifacts (docs, decision records, source): those belong under version control.
    0 installs
  40. ▌
    Tdd · spencermarx bundle
    Test-driven development, the red-green loop that produces tests worth keeping. Use whenever the user wants to build a feature or fix a bug test-first, mentions TDD or red-green-refactor, asks for integration or behaviour tests, or asks where a test should go or why a test is brittle. Do NOT use to root-cause a failure whose cause is unknown (`first-principles-investigation`), to review a finished diff (`ocr-review-loop`), or for a throwaway demo (`prototype`).
    0 installs
  41. ▌
    Ocr 3 · spencermarx bundle
    AI-powered multi-agent code review. Simulates a team of Principal Engineers reviewing code from different perspectives. Use when asked to review code, check a PR, analyze changes, or perform code review.
    8 installs
  42. ▌
    Ocr 2 · spencermarx bundle
    AI-powered multi-agent code review. Simulates a team of Principal Engineers reviewing code from different perspectives. Use when asked to review code, check a PR, analyze changes, or perform code review.
    5 installs
  43. ▌
    Demo 2 · spencermarx
    A demo skill loaded from the example's project directory. Replies with a one-line greeting.
    4 installs
  44. ▌
    Ocr · spencermarx bundle
    AI-powered multi-agent code review. Simulates a team of Principal Engineers reviewing code from different perspectives. Use when asked to review code, check a PR, analyze changes, or perform code review.
    1 install
  45. ▌
    Ocr Review To Approval Loop · spencermarx bundle
    Drive a PR to an approved code review by looping OCR's multi-agent review and address steps. Runs /ocr:review then /ocr:address repeatedly until the review verdict is APPROVE, then one final /ocr:address for leftover suggestions, posting every review and every address round to the GitHub PR as comments. Use when the user asks to 'review and address in a loop', 'iterate review until approved', 'auto review-and-fix this PR', 'loop /ocr:review and /ocr:address', or to take a changeset all the way to a clean APPROVE with the default reviewer team. Wraps the /ocr:review (.ocr/commands/review.md) and /ocr:address (.ocr/commands/address.md) skills; needs a feature branch with an open GitHub PR and the gh CLI for posting.
    7 installs
  46. ▌
    Swarm Orchestration · spencermarx
    Orchestrate multi-agent swarms with agentic-flow for parallel task execution, dynamic topology, and intelligent coordination. Use when scaling beyond single agents, implementing complex workflows, or building distributed AI systems.
    3 installs
  47. ▌
    Reasoningbank Intelligence · spencermarx
    Implement adaptive learning with ReasoningBank for pattern recognition, strategy optimization, and continuous improvement. Use when building self-learning agents, optimizing workflows, or implementing meta-cognitive systems.
    3 installs