Borrowing it
Nothing to install: this file belongs to Nizoka/pdfnative-mcp. 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/Nizoka/pdfnative-mcp/main/.github/instructions/mcp-server.instructions.mdgit clone --depth 1 https://github.com/Nizoka/pdfnative-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/instructions/nizoka/pdfnative-mcp/mcp-server)<a href="https://agentmods.dev/instructions/nizoka/pdfnative-mcp/mcp-server"><img src="https://agentmods.dev/badge/instructions/nizoka/pdfnative-mcp/mcp-server/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/instructions/nizoka/pdfnative-mcp/mcp-server"><img src="https://agentmods.dev/badge/instructions/nizoka/pdfnative-mcp/mcp-server.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 | $0.01433 | $0.01433 |
| Opus 5.5 | $0.00573 | $0.00573 |
| Sonnet 5.5 | $0.00287 | $0.00287 |
| Haiku 4.5 | $0.00143 | $0.00143 |
Grade A, and why
pdfnative-mcp mcp-server.instructions.md 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 1mo 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 — 44 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MCP Server Standards
Transport and lifecycle
- stdio transport by default; Streamable HTTP when
PDFNATIVE_MCP_PORTis set (loopback bind + Host/Origin guard; opt-in bearer token viaPDFNATIVE_MCP_HTTP_TOKEN, seesrc/auth.ts). - SDK:
@modelcontextprotocol/server^2.0.0 (MCP 2026-07-28, automatic 2025-era fallback). - Gracefully close the server on SIGINT/SIGTERM; stdin EOF ends the stdio server with exit 0.
- stdout is the JSON-RPC channel. Never
console.loginsrc/— logs go to stderr. The gate'ssmokeanddist-probesteps fail on a stray stdout line.
Tool contract
- Each tool module exports:
TOOL_NAME,TOOL_INPUT_SCHEMA, optionalTOOL_OUTPUT_SCHEMAhandler(args: unknown)
- Keep JSON schema and Zod validation aligned; Zod schemas are
.strict()(unknown top-level or nested keys →VALIDATION_ERROR, matchingadditionalProperties: false). Build both from the same constants where you can (src/color.ts,src/typography.ts,src/pdfx.ts). - Decode
pdfBase64/ DER / ICC inputs throughsrc/base64.ts(data: URI tolerated; PEM-vs-DER, double-encoding and empty-payload hints).Buffer.from(x, 'base64')never throws — do not rely on it to validate. - Return tool-level errors as
isErrorresponses with actionable messages — never throw across the MCP boundary. The one protocol-level exception: an unknown tool name ontools/callis a JSON-RPC-32602error with message[UNKNOWN_TOOL] Unknown tool: <name>. - Every failure carries a
ToolErrorcode. pdfnative parses lazily, so a damaged PDF can throw from any accessor afteropenPdf();classifyUnexpected()insrc/server.tsreports whatever escapes a handler asPDF_PARSE_FAILED. Map the failures you expect yourself; the net is the last line, not the first. - Every tool registered in
src/server.tsmust ship_meta.apiVersion(currentTOOL_API_VERSION, seedocs/API_STABILITY.md) and one or two executable_meta.examples(they are validated againstinputSchemain tests; more examples belong inexamples/*.json). tools/liststructure is pinned bytests/_fixtures/tool-shape.json(tests/catalogue-parity.test.ts, gate stepverify:tool-shape). Wording is free; any change to types, enums, defaults, constraints,required,additionalPropertiesor example count needsnpm run build && npx tsx scripts/tool-shape.ts --writeand a review underdocs/API_STABILITY.md§5.tests/catalogue-superset.test.tsadditionally rejects any removal / narrowing against the frozen 1.5.0 fixture (tests/_fixtures/tool-shape.v1.5.0.json); widening a plain schema intoanyOfis accepted only when a member still accepts every published value.- No
$ref/$defsand no$schemakeyword in input schemas (host compatibility); shared fragments are therefore inlined in every tool — keep their descriptions terse (tools/list≈ 306 kB, budgeted at 320 KiB bytool-shape.ts --check) and put the long form in a prompt or a guide. - Update
SERVER_INSTRUCTIONS(decision tree + pitfalls) and the relevant prompt whenever you add a tool or change its behaviour.
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.
- 1mo ago Changed · +8 lines · +538 tokens per session da81344a6e8a
- 1mo ago First seen · 36 lines · 895 tokens per session scan A 72032e24880a
pdfnative-mcp mcp-server.instructions.md is an instructions file published in the GitHub repository Nizoka/pdfnative-mcp (2 stars, last pushed 7d ago), licensed MIT. It adds 1,433 tokens to every session, about $0.0057 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 2026-08-31.
Other instructions, from other repositories
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).
vscode model-prompts.instructions.md
Model-specific prompt authoring and registry guidelines.