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 pvliesdonk/markdown-vault-mcp --skill researching-referencesgit clone --depth 1 https://github.com/pvliesdonk/markdown-vault-mcpWrote 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/pvliesdonk/markdown-vault-mcp/researching-references)<a href="https://agentmods.dev/skills/pvliesdonk/markdown-vault-mcp/researching-references"><img src="https://agentmods.dev/badge/skills/pvliesdonk/markdown-vault-mcp/researching-references/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/pvliesdonk/markdown-vault-mcp/researching-references"><img src="https://agentmods.dev/badge/skills/pvliesdonk/markdown-vault-mcp/researching-references.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 | $0.00099 | $0.02379 |
| Opus 5 | $0.00049 | $0.01189 |
| Sonnet 5 | $0.00020 | $0.00476 |
| Haiku 4.5 | $0.00010 | $0.00238 |
Grade A, and why
researching-references 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 2d 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 — 158 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Researching references
A reference is a page under docs/design/reference/ that records how
something outside this repository behaves: the rules of a markdown
dialect, what a git remote refuses and how it says so, the quirks of a file
format or a vendor API. It exists because the alternative is worse in a
specific way: an agent that meets such a question mid-task answers it from
parametric memory, which is consistently incomplete, outdated, or wrong in
the details that matter, and the design doc only ever records the facets that
already broke. Six issues about one rule set, each rediscovering one facet,
is the failure this skill exists to stop.
A reference is not the design doc. docs/design/ says what this project
does and why; a reference says what the world does, with the evidence. When
the two disagree on purpose, the reference records the world and links to the
design decision that departs from it.
When to write or refresh one
- A bug or feature turns on an external behaviour and no reference under
docs/design/reference/covers it. Write one before fixing: the fix is then a consequence of the reference, not a guess the reference later contradicts. - A reference exists but its
stale_afterdate has passed, itssubject_versionis no longer what the project targets, or itsstatusisdeprecatedwithout a successor. A stale reference is re-researched, never trusted and never silently extended. - A bug turns out to be rooted in a facet the reference lacks. The fix closes with a reference entry, not only a narrative in the design doc.
Do not write one for the project's own behaviour, for anything a single link to a stable spec section covers, or as a tutorial. Scope a reference to what the code depends on: a git reference for a module that only pushes and reads history covers refusals, identity, and credentials, not branching strategy.
What a reference looks like
What ships with it
1 file 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.
- 2d ago First seen · 158 lines · 99 tokens per session scan A a4df2b336638
researching-references is a skill published in the GitHub repository pvliesdonk/markdown-vault-mcp (32 stars, last pushed today), licensed MIT. It adds 99 tokens to every session and 2,379 once invoked, about $0.0005 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 2026-09-07.
Other skills, from other repositories
mnemonic
Search local indexed markdown knowledge bases. Use when the user asks to find notes, dig up a concept from personal docs, cross-reference ideas across wikis, or answer from indexed local files. Triggers on "look up in notes", "search my docs", "what did I write about", "find in my vault", "check my index", "retrieve…
youtube-fetcher
Retrieve YouTube transcripts and subtitles, summarize or analyze what was said, or save an Obsidian-ready Markdown knowledge-base note with captions, creator metadata, chapters, language, and source provenance. Use for a YouTube URL or video ID when the request needs spoken content or an archival note. A bare YouTube…
link-memory
Use after important user-approved decisions, when durable context should be proposed or reviewed, and for explicit Link memory lifecycle work: remember, recall, review, update, archive, restore, forget, or explain local memories through the CLI without requiring MCP.
link-retrieve
Use before answering work that may depend on user memory, project history, source-backed notes, or prior decisions; retrieve compact Link context through the CLI without loading the whole wiki or requiring MCP.
link-ingest
Use when raw files are present, source pages look stale, or a user asks to ingest notes into Link; refresh source-backed wiki pages, propose memories, and validate updates through the CLI without MCP.
reduce
Extract structured knowledge from source material. Comprehensive extraction is the default — every insight that serves the domain gets extracted. For domain-relevant sources, skip rate must be below 10%. Zero extraction from a domain-relevant source is a BUG. Triggers on "/reduce", "/reduce [file]", "extract…