Skip to content
← Tools / Skills
Skill · .md packagetderoussel/10 trustMetrics-checked
TD Agent Skills logo

TD Agent Skills

18 process skills and 4 subagents for verification gates, audits, and multi-agent briefs.

Visit official site →

About

A personal library of eighteen skills plus four subagents focused on process rather than code: a ship-gate that verifies before you claim something is done, a codebase auditor that separates proven findings from inferences, a decision council that runs choices through independent lenses, and an inbox triage that drafts replies without sending. It's distributed as a Claude Code plugin marketplace you add with one command. This is a solo-maintainer project with minimal adoption so far, so treat it as a set of interesting workflows to try rather than battle-tested infrastructure.

What it can do

  • Ship-gate verification before claiming completion
  • Codebase audits with evidence-backed findings
  • Decision council with blind peer review
  • Multi-agent work briefs with file ownership
  • Overnight build prompt composition
  • Inbox triage with draft-only replies
  • Session logging for durable handoffs

Setup

  1. In Claude Code, add the marketplace: /plugin marketplace add tderoussel/agent-skills
  2. Install: /plugin install agent-skills@agent-skills
  3. Skills load automatically when their description matches your task.
  4. Or invoke by name: "use ship-gate before you tell me that's merged."

Prerequisites

Claude Code with plugin support.

Use cases

codeagentsautomation

Works with

Claude Code

Trust score — 5/10

/10 trustMetrics-checked

Scored 5/10 (metrics tier, metrics-v1). Evidence: Repo exists with documented 18-skill structure and plugin marketplace install path; Only 1 GitHub star — minimal adoption, solo maintainer; Skill descriptions are unusually specific and well-written, which is a positive signal; Not hands-on tested; treat as experimental. License: Public repo; check license file before reuse.

★ 1 GitHub stars

How the trust score is built

Full methodology →
Adoption · stars/downloads, log-scaled vs peers
2.0/10
Maintenance · recency, releases, issue load
5.0/10
Openness · license + docs
8.0/10
Manual review · sanity read + red-flag check
5.0/10
  • · Repo exists with documented 18-skill structure and plugin marketplace install path
  • · Only 1 GitHub star — minimal adoption, solo maintainer
  • · Skill descriptions are unusually specific and well-written, which is a positive signal
  • · Not hands-on tested; treat as experimental

Sources

Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.