polycode-public
Official@polycode-public
polycode-public ships skills for plain-prose writing rules, playtest hardening loops, backlog dispatch, status synthesis, and news-card iteration.
Agent Skills by polycode-public
Showing 5 vetted skills indexed across 1 GitHub repositories.
polycode-publicplain-prose
Applies plain-English writing rules and removes LLM-voice tells from human-facing text.
polycode-publicplaytest-edge-hunt
Runs iterative six-area conversation playtests against tmct to find, fix, and ship edge-case fixes.
polycode-publicdo-next
Dispatches NEXT.md open items as worktree-isolated sub-agents and merges each landed track.
polycode-publicrefresh-status
Regenerates STATUS.md from committed benchmark reports, CI pipelines, plan docs, and README claims.
polycode-publicnews-feed-quality
Iterates news card generation cycles to close gaps against a hand-written target card.
Frequently Asked Questions About polycode-public
FAQPage SchemaWhat tasks can I run using polycode-public's skills?▼
You can enforce plain-prose writing rules on human-facing text, run six-area playtest hardening loops against tmct, dispatch NEXT.md backlog items as worktree-isolated sub-agents, resynthesize STATUS.md from committed reports and CI, and iterate news.html card quality through capture-fix-merge cycles.
Who are polycode-public's skills designed for?▼
They target operators and maintainers of the polycode repos who write docs and site copy, harden the six standing user-journey areas, work the NEXT.md backlog, and keep STATUS.md and news.html cards accurate after each landed report or pipeline run.
How are polycode-public's skills invoked in practice?▼
Each skill triggers on explicit operator phrases: load plain-prose before writing text, say 'run a playtest' with an iteration budget, say 'do next' or 'clear NEXT.md', ask to refresh STATUS.md after a landed report, or say 'run the news loop' with an iteration number.
What inputs and dependencies do these skills require?▼
refresh-status needs already-committed benchmark reports, PLAN docs, README claims, and the latest CI pipeline on main. do-next needs an existing NEXT.md with open items. playtest-edge-hunt needs the tmct conversation target, and news-feed-quality needs news.html fixtures and a live cycle.
Does refresh-status run benchmarks or modify source documents?▼
No. refresh-status only resynthesizes STATUS.md from already-committed benchmark reports, PLAN docs, README claims, and the latest CI pipeline on main. It explicitly runs no benchmarks and edits no source documents, making it safe to invoke after any landed report or status change.