dimitritholen
UserThe organiser for your AI. A backlog, a memory and a brief for your coding agent.
Categories
Indexed Skills (21)
land-tasqx-pr
Land a tasqx task's branch on protected main without waiting on it — push, open the PR, arm auto-merge (rebase, without --delete-branch) and move on to the next task; at the next task boundary check the PR, send failed checks or real Qodo findings back to a builder, and once it is MERGED reinstall, verify, clean up and close the task once. Use it when a task's branch has passed review under CLAUDE.md's "One task, one branch" rule, or when the user asks to land, merge, ship or open the PR for a task ("land task
retro
Retrospective on the tasqx tasks completed in this session — review the decisions and rework that got them done, record the lessons on each task, and decide whether a procedure earns a skill. Use at the end of a session, or when the user says retro, retrospective, post-mortem, lessons learned, or asks what we learned. Never triggered by completing a task.
tasqx-workflow
Use tasqx (MCP tools + CLI) for backlogs, task tracking and knowledge memory. Use it whenever the user starts or continues feature work, wants work planned or tracked, asks what is open or what to pick up next, wants a status report, or wants a decision or convention stored or looked up — even when they never say "tasqx". Skip it for one-off questions and one-line fixes.
spec
Turn the current conversation into a spec — no interview, just synthesis of what you've already discussed. Use on /1337:spec, or when the user asks to spec out, write up or formalise a feature you've been discussing. Writes only the spec, never code.
proudhuman
Tell the story of a pull request or branch from the Claude Code transcripts that produced it, with the human's steering as the subject. Use when the user asks what they actually did on a PR, who is accountable for a change, whether a PR was reviewed by a human, how much of a branch the AI wrote, for a story or self-review of their AI-assisted work, or to run proudhuman.
handoff
Compact the current conversation into a handoff document for another agent to pick up. Use on /1337:handoff, or when the user asks to hand off, pick this up later, or continue in a new session.
implement
Implement a piece of work from a plan, a spec or a set of tickets, one task at a time with review between steps. Use on /1337:implement, or when the user wants a planned change built end to end.
assumptions
Audit a plan, spec, PRD, design doc or diff for unchecked assumptions that would flip the decision, each with what changes if it is wrong. Use on /1337:assumptions, or when the user asks what a plan assumes, whether a plan is safe to act on, or to pressure-test a design before building. Read-only.
audit
Audit a whole repo or module for over-engineering that is alive and in use — speculative generality, pass-through layers, frameworks where a function would do, homegrown versions of stdlib features — and hand back a delete-list with evidence. Use on /1337:audit, or when the user asks whether the codebase is over-built, what can be simplified repo-wide, or says a codebase feels bloated or over-abstracted. Read-only; never edits. Diff-scoped review → /1337:review; dead code and markers → /1337:debt.
debt
Sweep a repo or directory for accumulated debt �� TODO/FIXME/HACK markers, dead code, obsolete workarounds, stale dependencies — and hand back a delete-ledger ordered by payoff. Use on /1337:debt, or when the user asks what can be deleted, what is dead, to clean up, pay down debt or find quick wins. Read-only; never edits.
scope
Shrink a feature request to the smallest version that still achieves its goal before any code is written. Use on /1337:scope, or when the user asks to trim, cut down, scope or descope a feature, or says "what is the minimum", "smallest version", "MVP of" or "do we need all that". Read-only; never builds.
tier
Split a task into steps and assign the cheapest model tier that can do each one, for dispatching to 1337:builder or planning any multi-step change. Use on /1337:tier, or when the user asks which model to use, how to break a task into steps, how to dispatch to subagents, or says a task is too big for one go. Read-only.
visual
Make image, SVG, video or speech files through an OpenRouter model picked from a Jev-ranked, priced list, and store the OpenRouter key once. Use on /1337:visual, "/1337:visual setup", when the user asks to generate, draw, render or narrate something as a file, when a prompt carries a "[1337 visual]" context block from the hook, or when a Jev-backed script reports a missing key.
seed-corpus
Turn this shell test suite into a replayable corpus, or seed a corpus from tests/<x>.test.sh with pending rows for the known gaps. Use when you need to capture and replay all the checks a suite exercises as data-driven test rows, so later changes can verify against exact payloads instead of reimplementing the original logic.
diagnose
Diagnosis loop for hard bugs, regressions, flaky failures, and performance problems. Use on /1337:diagnose, or when the user says "diagnose"/"debug this", reports something broken/throwing/failing/slow, hits a regression, a test is flaky, or asks "why does X".
domain-modeling
Build and sharpen a project's domain model. Use on /1337:domain-modeling, or when discussing codebase terminology, writing or editing a CONTEXT.md, or recording or editing an ADR.
grill
Grill the user relentlessly about a plan, decision, or idea, sharpening the domain model and recording settled decisions as it goes. Use on /1337:grill, or when the user wants to stress-test their thinking, or uses any 'grill' trigger phrases.
tdd
Use on /1337:tdd, or when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests. Edits test files and code.
unslop
Use on /1337:unslop, or when the user says unslop, when prose reads AI-generated, or as an editing pass after drafting docs, articles, or PR descriptions. Edits text files and prose.
terse
Report the terse mode level that the verbosity governor enforces (off, on, hard) and where it is set. Use on /1337:terse, or when the user says replies are too long or chatty, to stop explaining so much, be terser, or keep answers short from now on.
merge-conflicts
Use on /1337:merge-conflicts, or when you need to resolve an in-progress git merge/rebase conflict. Edits conflicting files.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.