Content Run — Data skill for Claude Code
Run the Harbourlight pipeline once - research, draft, deterministic check, independent review, then a human decision that is logged.
How to install Content Run
Installs to ~/.claude/skills/cla1redonald-harbourlight-content-run/SKILL.md
mkdir -p ~/.claude/skills/cla1redonald-harbourlight-content-run && curl -fsSL https://raw.githubusercontent.com/cla1redonald/harbourlight/HEAD/commands/content-run.md -o ~/.claude/skills/cla1redonald-harbourlight-content-run/SKILL.md Restart Claude Code, or start a new session, for it to be picked up.
What Content Run does
description: Run the Harbourlight pipeline once - research, draft, deterministic check, independent review, then a human decision that is logged. argument-hint: "[story id, e.g. S02, to skip research]"
Content run
Run the pipeline once, end to end. Read `CLAUDE.md` first if you have not already; its hard rules apply to every step. Never publish anything: this command ends with files on disk and a row in the audit log. A person publishes.
Note the current `voice_version` from `refere
Alternatives in Data
- Clinical Decision Support — Evidence-based clinical decision support tools 6.4k ★
- Ars Full — ARS full pipeline — research → write → review → revise → finalize 4.5k ★
- Ralph Marketer Status — Check the current status of the content pipeline and Ralph's progress 698 ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
Reframe Orchestrator
Reframe a repo's .claude/orchestrator.md from a human-paused dispatch persona into a mode-independent invarian
Ckw Skills
ckw-skills — Conner K. Ward's Claude Code marketplace of skills + MCP servers: human-in-the-loop studios, dete
Durable Claude Agent
Claude text-to-SQL analyst over a governed lakehouse, run as a Temporal workflow: human-approval gate, determi
Conversation Reply
Investigate a PostHog Conversations support ticket - find root cause via git/code/PostHog data, verify whether
Lecture Notes Pipeline
Records lectures on a Mac, transcribes them locally with whisper.cpp while the class runs, and turns them into
Self Eval
Self-improvement loop. After a pipeline run, reviews how faithfully it executed step-by-step (from the determi
Related Agents
Blameless Critic
Use when a drafted incident postmortem needs an independent review against a written blameless standard before
Absorption Check
Warm per-round absorption check for design.md and track files: two-way coverage match between the research log
Linker
Linker for the thruuu article writing pipeline. Receives the humanized draft, the Links element from the brief