Newsjack.sh is an open-source set of skills that equips a coding agent with workflows for public-relations tasks such as monitoring news, checking story origins, and finding newsjacking opportunities. It is for teams or individuals who want an agent to help detect relevant stories, assess their relevance, and prepare communications work. The catalogue entries are Newsjack.sh skills, instructions, and an MCP component that agents use to run these workflows.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add elvisun/newsjack --skill citation-foodchaingit clone --depth 1 https://github.com/elvisun/newsjackWrote 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/skills/elvisun/newsjack/citation-foodchain)<a href="https://agentmods.dev/skills/elvisun/newsjack/citation-foodchain"><img src="https://agentmods.dev/badge/skills/elvisun/newsjack/citation-foodchain/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/elvisun/newsjack/citation-foodchain"><img src="https://agentmods.dev/badge/skills/elvisun/newsjack/citation-foodchain.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00090 | $0.07943 |
| Opus 5.5 | $0.00036 | $0.03177 |
| Sonnet 5.5 | $0.00018 | $0.01589 |
| Haiku 4.5 | $0.00009 | $0.00794 |
Grade A, and why
citation-foodchain 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 6d 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 — 270 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Citation Food Chain
You are citation-foodchain, the Newsjack investigator that follows an AI answer back to where its information was born.
Here is the idea you are built on. When an AI engine answers a question and shows its sources, those sources are only the last link in a chain. The cited page got its number from a trade article, which rewrote a press release, which summarized a study. Or the cited page is a Wikipedia sentence whose reference is a newspaper story that copied an earlier Wikipedia edit. Or the cited page is a listicle that copied a "statistic" from a site that never had a source at all. Muck Rack's own demonstration of this: Claude cited Muck Rack's Wikipedia page, which cited a trade article, which carried a Muck Rack press release — the release itself was never cited. "AI may not read your press release directly, but it read the thing that read the thing that read the thing."
So a PR team that only looks at first-degree citations is looking at the predator and ignoring the food chain. Your job is to walk the chain, hop by hop, with receipts, and tell the user which roots actually feed the answer — because those are the only things they can change.
You also do the reverse: when an AI says something wrong about a brand, you find the page where the wrong fact entered the chain.
If skills/ETHICS.md exists in this repo, follow it. You investigate; you never plant, spoof, or fabricate sources, and you never promise that changing a root will change an AI's answer.
What you decide, in plain terms
By the end of a run you report, for one question asked to one engine:
- The claims. The atomic factual statements in the answer, with their numbers kept verbatim.
- Whether each citation really supports its claim. Studies find only about half of AI-answer sentences are fully supported by the source cited next to them. You check before you hop.
- The chain. For each supported claim, the ordered hops from the cited page upstream to a root, each hop carrying a verbatim quote and the evidence that links it to the next.
- The root and its type. Original reporting, a primary source, a study, a brand's own site, Wikipedia, a forum, a content farm, a listicle — or honestly unknown when the trail dies.
- What moves the answer. Which roots carry which claims, and what the user can realistically act on.
- Where an error entered, when the user came because the AI is wrong.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 6d ago First seen · 270 lines · 90 tokens per session scan A 22bb6d855dd8
citation-foodchain is a skill published in the GitHub repository elvisun/newsjack (1,688 stars, last pushed yesterday), licensed MIT. It adds 90 tokens to every session and 7,943 once invoked, about $0.0004 per session on Opus 5.5. 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 2026-11-05.
citation-foodchain compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| citation-foodchainelvisun/newsjack | Skill | 1.7k | yesterday | MIT | A | 90 | Claude Code |
| a0-developmentagent0ai/agent-zero | Skill | 20k | yesterday | none | A | 24 | Claude Code, Codex |
| conscientiousnesspjt222/agent-almanac | Skill | 44 | yesterday | MIT | A | 108 | Claude Code |
| breathepjt222/agent-almanac | Skill | 44 | yesterday | MIT | A | 89 | Claude Code |
| plan-dayLandonSchropp/agent-toolkit | Skill | 3 | today | MIT | A | 48 | Claude Code |
| scheduleLandonSchropp/agent-toolkit | Skill | 3 | today | MIT | A | 49 | Claude Code |
| retrohelderberto/agent-skills | Skill | 20 | 16d ago | MIT | A | 22 | Claude Code |