Borrowing it
Nothing to install: this file belongs to TheSmokeDev/taskchad-os. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/TheSmokeDev/taskchad-os/master/.claude/skills/ai-citation-authority-wave/assets/archon/commands/citation-authority-write.mdgit clone --depth 1 https://github.com/TheSmokeDev/taskchad-osWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/thesmokedev/taskchad-os/citation-authority-write)<a href="https://agentmods.dev/commands/thesmokedev/taskchad-os/citation-authority-write"><img src="https://agentmods.dev/badge/commands/thesmokedev/taskchad-os/citation-authority-write.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00014 | $0.00865 |
| Opus 5 | $0.00007 | $0.00432 |
| Sonnet 5 | $0.00003 | $0.00173 |
| Haiku 4.5 | $0.00001 | $0.00086 |
Grade A, and why
citation-authority-write scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 105 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Write exactly one queued page
This command is executed with fresh context for each page. Write one page, validate it, repair that same page if necessary, then stop. Never advance to a second queue item in the same iteration. Do not use a Reddit account, run git, publish, deploy, or request indexing.
Find the one target
Run:
python .claude/skills/ai-citation-authority-wave/scripts/citation_authority.py next-page
If it reports complete: true, output <promise>COMPLETE</promise> and stop.
Otherwise read only the returned target plus:
.citation-authority/site-profile.json.citation-authority/candidate-plan.json- the configured content reference file or frontmatter template
- source URLs and existing pages needed to avoid duplication
Do not alter the approved query, title, slug, route, output path, opening sentences, H2 list, sources, internal links, or brand-role passage.
Write the draft and integrated page
Create both:
.citation-authority/pages/<target-id>.draft.md- the target's approved
output_path
The two files must have the same normalized content. Match the site's real frontmatter and content conventions. Keep content server-renderable without client-side reveal gates.
Content rules:
- The first paragraph contains the two approved direct-answer sentences verbatim and in order.
- Use exactly the approved H2 headings, in order. Build a distinct argument under each rather than filling a template.
- Meet
quality.min_main_words, but remove repetition and keyword-density filler. The page must deserve to exist without search traffic. - Cite every approved external source naturally. Link every approved internal route naturally, including a service hub.
- Include the approved transparent brand-role passage verbatim.
- Synthesize Reddit discussion honestly. Do not imitate comments, invent consensus, imply endorsement, or automate on-platform activity.
- On regulated sites, distinguish general information from licensed or legal advice. Every material number and regulated factual claim needs the approved authoritative source.
- Spanish pages are fully native Spanish, including labels, dates, CTA, and metadata. They are not translations and do not use English scaffold copy.
- Do not invent prices, rates, savings, approval, eligibility, rankings, customer counts, case studies, quotes, or outcomes.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 8d ago First seen · 105 lines · 14 tokens per session scan A c9048ed2fd7b
citation-authority-write is a command published in the GitHub repository TheSmokeDev/taskchad-os (23 stars, last pushed yesterday), licensed MIT. It adds 14 tokens to every session and 865 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
cierre
A sales call just ended: turn its transcript into the full follow-up (CRM, tasks, email draft, reminder, coaching).
daily-journal
Daily journal interview and entry creator with emotional floor tagging.
deconstruct
First-principles analyst: surface hidden assumptions, find foundational truths, rebuild from scratch.
diagnose
Run a self-check on your AI Brain Starter install (CLAUDE.md, Meta folder, skills, hooks, MCPs).
evolve
Cluster hardened instincts into a proposed Command / Skill / Agent (Instinct Engine).
graphify
Turn your vault into a clustered knowledge graph with HTML and JSON outputs.