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 Chael-Chael/dsh-reference-anything --skill reference-source-authorgit clone --depth 1 https://github.com/Chael-Chael/dsh-reference-anythingWrote 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/chael-chael/dsh-reference-anything/reference-source-author)<a href="https://agentmods.dev/skills/chael-chael/dsh-reference-anything/reference-source-author"><img src="https://agentmods.dev/badge/skills/chael-chael/dsh-reference-anything/reference-source-author/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/chael-chael/dsh-reference-anything/reference-source-author"><img src="https://agentmods.dev/badge/skills/chael-chael/dsh-reference-anything/reference-source-author.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 | $0.00085 | $0.02084 |
| Opus 5.5 | $0.00034 | $0.00834 |
| Sonnet 5.5 | $0.00017 | $0.00417 |
| Haiku 4.5 | $0.00009 | $0.00208 |
Grade A, and why
reference-source-author scanned grade A with 1 finding 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 yesterday.
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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- Optional children: `children: [{ key, title, kind: 'image' | 'file' | 'text', item? }]` plus `fetch(id, child, maxBytes)` returning bytes; the core saves the file (cap from `files.maxBytes`) and adds pixels for images How it starts
The opening of the file, as written. The whole thing — 115 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Write a Reference Anything source
The user describes what they want to @. You deliver a working DSH plugin in their profile that:
- registers a Host source plugin the model reaches through the single
referencetool (list,search,read; images and files are children read withas: "image"oras: "file"). WithdefineReferenceSourceyou write onlyloadand usuallylist; you write no tools, and - usually adds an
@menu group so the user can pick items in the input box.
There is no fixed list of allowed sources. Read what the user asked for, from wherever it lives: another installed plugin's services, Remotes or browser components, a web page or API, files, a database, a command's output. Build exactly that.
Environment
- DSH profile:
{{profileName}}, directory{{profileDir}} - DSH version:
{{dshVersion}}(use it for every@deepseek-ai/*package you install) - Reference Anything version:
{{raVersion}} - Reference Anything package directory:
{{raDir}}(itslib/reference-source.d.tsandlib/sdk.d.tsare the authoritative types)
1. Pin down the source
Settle these before writing code. Ask the user only about what you cannot work out yourself, in one short message:
- Items: what one item is, its stable id (must survive restarts without a lookup table; never a credential, a user name or an absolute path; never a path the user could escape with
..), and its title. - Origin: where the list and the body come from, and whether reading needs network, a login or a running process.
- Reading: what
loadreturns: the whole object as records (text blocks, chat turns, rows; oldest first), plus children such as attachments. - Menu: whether the user wants an
@group (default yes) and a shortcut prefix such asnote:.
To reuse another plugin, find its package under {{profileDir}}/node_modules and read its package.json exports and lib/**/*.d.ts. Host services are reached with inject: ['theirService'] and ctx.theirService. Browser services are reached with ctx.reflect.get('remote.<service>') or the client service they register.
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.
- yesterday First seen · 115 lines · 85 tokens per session scan A 106bb745f175
reference-source-author is a skill published in the GitHub repository Chael-Chael/dsh-reference-anything (27 stars, last pushed 2d ago), licensed MIT. It adds 85 tokens to every session and 2,084 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-11-15.
reference-source-author compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| reference-source-authorChael-Chael/dsh-reference-anything | Skill | 27 | 2d ago | MIT | A | 85 | Claude Code, Codex |
| rea-changelog-updatemorluto/rea | Skill | 462k | today | MIT | A | 33 | Claude Code |
| feynman-perspectivealchaincyf/nuwa-skill | Skill | 38k | yesterday | MIT | A | 212 | Claude Code, Codex |
| x-mastery-mentoralchaincyf/nuwa-skill | Skill | 38k | yesterday | MIT | A | 197 | Claude Code, Codex |
| elon-musk-perspectivealchaincyf/nuwa-skill | Skill | 38k | yesterday | MIT | A | 228 | Claude Code, Codex |
| ilya-sutskever-perspectivealchaincyf/nuwa-skill | Skill | 38k | yesterday | MIT | A | 183 | Claude Code, Codex |
| novel-reviewerLiPu-jpg/Openwrite | Skill | 960 | yesterday | Apache-2.0 | A | 49 | Claude Code, Codex |