Spill — Development skill for Claude Code
Writes an evidence-backed PRD or issue draft (\"spill the tea\").
How to install Spill
Installs to ~/.claude/skills/nosy-hq-nosy-spill/SKILL.md
mkdir -p ~/.claude/skills/nosy-hq-nosy-spill && curl -fsSL https://raw.githubusercontent.com/nosy-hq/nosy/HEAD/commands/spill.md -o ~/.claude/skills/nosy-hq-nosy-spill/SKILL.md Restart Claude Code, or start a new session, for it to be picked up.
What Spill does
description: "Writes an evidence-backed PRD or issue draft (\"spill the tea\"). Never sent out." argument-hint:
Use the Nosy skill (`${CLAUDE_PLUGIN_ROOT}/skill/SKILL.md`, rules: `${CLAUDE_PLUGIN_ROOT}/skill/rules.md`). Wherever the command docs say ``, it means `${CLAUDE_PLUGIN_ROOT}/skill` here.
Follow the `spill ` steps in `${CLAUDE_PLUGIN_ROOT}/skill/commands/spill.md` from start to finish.
Topic: $ARGUMENTS
Alternatives in Development
- Tech Debt — Multi-agent technical-debt assessment — fans out region-scoped + specialist hunters, verifies findings, scores 107 ★
- Diagnose Rota Routing — Explain why the rota gateway sent a request to a particular claude silo by reconstructing the scoring decision 67 ★
- Accuracy Comment — Accuracy-scout step 5 — draft accuracy comment, update issue, add label 27 ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
Cr Tune
Mine the CR ledger for disproved classes per critic and draft citation-backed tuning proposals (.coderabbit.ya
Network Outreach
Identify 2 to 3 strategic humans at a target company, draft a research-backed peer-to-peer outreach message, a
Define Issue
Define a new tracker issue (or project with multiple issues) through a conversation. Explores the codebase, in
Ad Fatigue
Handle an ad whose creative is wearing out. The designer builds variants that change one element each, the sch
Spread
Work out where a finished project goes and what each place demands of a newcomer — drafted posts, none of them
To PRD
Turn the current conversation context into a PRD and publish it to the project issue tracker. Use when user wa
Related Agents
Legal Review Lens
Read-only independent reviewer for one bounded legal-draft lens. Reports evidence-backed findings and never ed
Cloud Harness
Evidence-backed comparison of where the system should run (cloud and agent harness). Writes docs/architecture/
Loa Scout
Read-only evidence gathering and fan-out search for Loa skills. Never verdict-bearing review, never writes.