Shared core for windowsill plugins: assistant state store with atomic writes, remind-once etiquette, and schema-versioned JSON persistence. One file per plugin in the user's state directory.
0 on GitHubby Saharv0.2.0MITupdated 29 days agoutility
/
Plugin sill-core from marketplace windowsill in saharkit/windowsill. Commands read from the repo's manifest — 29 days ago.
What it installs
No skills, commands, agents, hooks or MCP servers were found in the standard plugin folders.
Two-way voice for Claude Code with pluggable speech providers: a config registry picks OpenAI, ElevenLabs or Deepgram per direction, or keeps every clip local/self-hosted (faster-whisper STT; Silero, XTTS-v2 or Ukrainian TTS on the bundled server) — switching backends is a config entry, never a code change. A Stop hook speaks the lines Claude marks, a push-to-talk hotkey dictates into your prompt, and /voice-setup installs it all. Requires sill-core (install both: /plugin install voice-loop@windowsill && /plugin install sill-core@windowsill).
One line at the bottom of Claude Code: how much of the context window is gone, how much of the weekly budget is spent, and which model is answering. Every figure is read from the payload the harness pipes to the script on stdin, so nothing is host-specific. Needs node and nothing else.
A handbook you hand to your agent: material for the general judgement calls — architecture decisions, planning, the rules that hold whatever is being built. Nothing here is tied to a language, a stack or a framework, and nothing installs a runtime, calls a network or needs hardware. It holds two kinds of thing: an instrument you reach for at a moment — `/target-world-walk`, which derives what a plan needs by replaying one concrete unit of work through the system you intend to have — and two standing rules that are never invoked at all — the flagellant rule (which reports a mistake as fact plus fix rather than apology, stops when the stopping condition is met instead of re-checking finished work, spends care where an error cannot be taken back, and reports completion only by first naming what would still be outstanding if it were wrong) and the non-stop rule (which says: do the work and report it rather than asking permission, write in the past tense, never hand back the turn while work remains, ask a question only where the work becomes pointless or unsafe without the answer, probe your own rights with a command before saying a task needs a human, report done only with proof, and never write farewell summaries). Both standing rules arrive through a SessionStart hook, so they compose with whatever output style you have chosen rather than replacing it.
Editorial diagrams in your brand, as self-contained HTML and SVG. Architecture, flowcharts, sequence, state machines, ER and database schemas, timelines, swimlanes, org charts, Gantt, Wardley maps, user journeys, service blueprints, UML class, lifecycle phase maps, and charts from bar, line and scatter to Sankey, treemap and heatmap, plus many more. Redraws draw.io, Mermaid, Excalidraw and PlantUML files. Exports SVG and PNG.
QuickBooks business insights and actions: financial health briefings across P&L, cash flow, balance sheet and A/R, industry benchmarking, payroll cost analysis, overdue-invoice reminders, estimate/invoice drafting from email, payroll onboarding and base-pay changes (with confirmation), and QuickBooks Capital lending guidance. Requires the Intuit QuickBooks connector.
Error analysis, trace debugging, and multi-agent problem diagnosis
3 commands · 2 agents
wshobson/agents 42,461
Maintain sill-core?
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.