market
claudeMarket research and go-to-market planning for business ideas.
2 skills
Skills wanted in every project.
Plugin core from marketplace agentic-coding in magroski/agentic-coding. Commands read from the repo's manifest — today.
epicPlan product work before any task is planned - a roadmap of epics or a single epic. Interviews the founder to find the real outcome, settles every business rule, success metric, limitation and scope cut as an explicit decision, writes a product-only epic doc whose tasks fan out in parallel after one
file-issueFile a task as a GitHub issue written to one pattern - a product-level title, then What, Why, How, Definition of done, and When for dated work. Use when the user invokes it or asks to file, open, create, log or write up an issue, ticket or task, or to turn a finding, bug, plan step or TODO into one,
nextStart the day on the current repo - read milestones, issues, PRs, due checks, plans in progress and untriaged follow-ups, run the checks that are due, and brief the founder in one screen on what needs them, what is moving and what is ready, grouped into parallel lanes. On the founder's go, act as te
plan-tasksPlan every filed task of an approved epic before any is built - one task-skill planner per task in parallel, their interviews batched into one message to the founder, then a fresh cross-review of the plans against each other and the epic doc (business rules, exit criteria, acceptance criteria, share
setupSet up the current repository for agent work - GitHub issues in six types (feature, chore, check, research, bug, founder) with labels, issue forms and conventions, plus a Delivery section in the repo's agent instructions recording how it is tested, deployed, rolled back, observed and measured, which
taskPlan one task technically before it is built - read the issue, its epic's business rules, the code, tests, infra and data; interview the founder on what the task must achieve; choose the smallest approach with the ladder; turn the acceptance criteria into checkable criteria, a test map, performance
cross-reviewerReviews an epic's task plans side by side, against each other and the epic doc, before any is built - every business rule, exit criterion and acceptance criterion covered, shared interfaces that match, order respected, nothing out of scope. Spawned by the plan-tasks skill; changes nothing.
map-auditorAudits a task plan's Test map before any test is written - every criterion covered, the named wrong builds caught, and every row worth writing under the test rules. Spawned by the task skill; changes nothing.
reviewerChecks a task's criteria, then reviews its diff adversarially - correctness, risk, scale, missing tests, speed and lean - separating what this diff got wrong, which the task fixes, from the few problems it did not cause that are worth the founder's time. Spawned by the build skill; changes nothing.
test-auditorGates tests written before implementation - checks each test is the Test map row it claims to be, fails for its stated reason, and is sound. Spawned by the build skill; changes nothing.
test-writerWrites a task plan's Test map as failing tests before any implementation exists. Spawned by the build skill with a plan path; not for general test writing.
verifierIndependently checks finished work against pass/fail criteria it did not write, running each check itself. Spawned by the next skill to close out an epic, with the criteria and the output location; changes nothing.
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.