Skip to content
SkillModel-led3,493 tokens in use

dbs-chatroom

Runs a simulated panel of thinkers on your topic, with each expert speaking from their own framework and a judge summing up the clashes and next steps.

11,518 stars on the repository Custom / Unknowndontbesilent2025/dbskill · skills/dbs-chatroom
npx -y skills add dontbesilent2025/dbskill -g --all
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing dbs-chatroom costs

67tokens idle
3,493tokens in use

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Counts are approximate, made offline with cl100k_base.

Model dependence Model-led

Script-ledMixedModel-led

Mostly instructions, so how well it works depends on the model that follows them.

This describes the skill's structure, how much of it is scripts. It does not say how well it works.

Touches

Nothing detected: no shell, network, file writes, git or credentials in its files.

Declared tags come from the skill's frontmatter; detected ones from its scripts, its code blocks and the commands it names in inline code. This is detected, not audited.

How it scores

Quality

How well the skill is written, counted from its own files: its frontmatter, its steps, examples, stated output and constraints, its size and whether every file it ships is one it names.How we score it →
Fair
  • Frontmatter8/15
  • Structure16/25
  • Examples9/15
  • Output0/15
  • Size15/15
  • Resources0/15

Security

Starts at 100 and loses points for each thing in the skill's files that a reader should look at first.How we score it →
Excellent

No findings.

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Excellent
  • Development activity96
  • Maintenance100
  • Community64
  • Maturity28
  • Momentum40

dbs-chatroom turns a topic into a multi-voice discussion. You name the experts, or let the Skill propose three to five whose theories cover different parts of your question, and you confirm the list. Each expert then answers in about 200 words, and a judge closes the round with a quality check, missed angles and two or three actions.

The room stays open for follow-ups, objections and new questions until you end it. Every first round opens with a notice that the speeches are simulated from public work and are not the person’s own words.

What You Get

  • A proposed panel of three to five experts, each with one line on their framework and role, before anything starts
  • An independent answer of about 200 words from each expert
  • A judge summary covering discussion quality, blind spots and two or three actions
  • A simulation notice at the top of the first round
  • Further rounds on your objections and new questions with the same panel

Common Use Cases

  • Testing a business or life decision against several competing theories
  • Naming your own panel, such as an investor, a psychologist and an economist, for a topic
  • Pushing back on a conclusion and letting the panel revise it
  • Finding the angle nobody in the first round raised

What Makes It Unique

The panel is chosen by theory, not fame. A candidate needs a clear core claim, a record of long testing and criticism, and a distinct job in your discussion, and content-style commentators are excluded from suggestions. Experts answer independently so one does not echo another, and the judge step is treated as the core value of the Skill.

Under The Hood

The agent reads SKILL.md and follows it. When the host supports sub-agents, such as the Agent tool in Claude Code, it launches one sub-agent per expert in parallel with a prompt it writes for each; otherwise it writes the speeches one after another without reading earlier ones. The Skill has no scripts, runs no shell commands, calls no web services itself and writes no files.

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers